- Probably Approximately Correct: Nature’s Algorithms for Learning and Prospering in a Complex World by Leslie Valiant
(New York: Basic Books, 2013), 208 pp.
Every era uses its most sophisticated technological device as its metaphor for how the universe works. Newton and his era used a watch with its exactly interlocking pieces. Our own era sees the universe working like a giant computer program. The laws of physics are, for us, like programs that tell the hardware how to act. In 2002, Stephen Wolfram published a controversial book arguing that a lot of biology and physics can be described in terms of cellular automata. In the book here under review, Leslie Valiant promotes the idea that large parts of biology and human nature can be understood as comprising a machine that learns algorithms. The highest award given in computer science is the Turing, which Valiant received in 2010. When a Turing laureate writes a book about ways of understanding processes in the world, one is well advised to take an interest. Algorithms are set rules about how to deal with inputs. Valiant shows that, often, an algorithm can be improved if it changes along with its environment. He calls his self-correcting algorithms “ecorithms” and shows that they are not only essential in describing evolution and learning but also in understanding “true” artificial intelligence. I suspect he has a point, and the point is made in a way understandable to the reader of popular science. In any case, there is no higher math here to scare anyone away. [End Page 340]
Noson S. Yanofsky is professor of computer and information sciences at Brooklyn College and the author of The Outer Limits of Reason: What Science, Mathematics, and Logic Cannot Tell Us and Quantum Computing for Computer Scientists. He coauthored Mathematics via Symmetry with Mark Zelcer.