The mind's mirror Risk and reward in the age of AI

MIT computer scientist Rus and science writer Mone follow up The Heart and the Chip with a rosy assessment of how artificial intelligence might change society. The authors suggest that AI will speed up advancements in medicine, describing how Canadian scientists synthesized a new cancer drug by using AI to first identify "weak spots" in the proteins that contribute to the disease and then to propose designs for drugs that could exploit those weak spots. AI can also provide one-on-one tutoring services, the authors contend, citing a study that found a program capable of adapting "its lessons in real time based on the participant's engagement" successfully helped narrow one elementary school class's achievement gap. Rus and Mone express confidence that AI will supplement rather than replace human workers, tackling mundane tasks so that people are free to focus on more creative activities. The authors' optimistic perspective serves as a refreshing corrective to AI doomsaying, and they offer practical suggestions for curbing what dangers the technology does pose, advocating for federal oversight of AI development and for disclosures to consumers when AI has been used to generate content. This enlightens. (Aug.)

An overview of the current and prospective benefits and perils of utilizing AI. Rus, a roboticist and the first female director of MIT's Computer Science and Artificial Intelligence Laboratory, and Mone follow up their first collaboration, The Heart and the Chip, by digging into the artificial brains inside the machines. The authors divide the book into three parts--"Powers," "Fundamentals," and "Stewardship"--and chapters bear concise, literal titles like "Speed," "Generating," and "Optimizing"; the longest is "Will AI Steal Your Job?" Accessible to a broad readership, that narrative delves into examples of how "we should think about making use of AI to accelerate processes in all areas of discovery, work, and life." These areas include science, manufacturing, logistics, inventory management, transportation, and other industries, as well as in the home and in artistic contexts. "We should all think about using AI as a tool," write the authors, "that can help us dig deeper, reach further, and imagine more boldly across all fields." They point to ways in which AI can help synthesize raw data into useful knowledge. "It's what we do with this information, and what ideas we generate from it, that really matters," the authors write. "What we really want is to turn this knowledge into insight." Citing numerous studies, they describe techniques for improving insight- and foresight-generating AI engines. Additionally, they present criticisms of the flaws, including vulnerability to hackers. One suggested protective measure is data distillation, which identifies common features of data and creates a new data set modeled on those patterns. To apply AI's potential to a wide audience, the authors conclude, repeatedly, that people are more necessary than ever. Though the authors work in a highly technical field, they do an excellent job of speaking to nontechnical audiences. This equitable work offers something for both AI enthusiasts and skeptics. Copyright (c) Kirkus Reviews, used with permission.