Understand, but not anticipate. To foresee but not to understand (Comprendre sans prevoir, prevoir sans comprendre)

In this book, the French scientist-popularizer of science, Hubert Krivin, talks about how ordinary science works, in which models are built on the basis of data that explain what is happening and have predictive power. And then he contrasts it with a modern approach to machine learning, where big data tracks correlations and creates predictions that are difficult to interpret and understand. Next, the author speculates whether fundamental science is necessary in the era of chatgpt (absent 5 Years ago, when the book was published.) He concludes that basic science is not going anywhere.

During the discussion, the author recalls the theory of deterministic chaos. (butterfly), analyzes why correlation is not causation, recalls Turing with his intelligence test and so on. In general, 150 There are many interesting examples and references to other areas of knowledge that are interesting to read.

#PopularScience #AI #Physics #Math