Reviewed the Book

Book review

Retrospect: The book "Why" examines the science of cause and effect Since Aristotle at the latest, the topic of causality has occupied the philosopher. Pearl has used the bow of his own carreer - first at RCA Laboratories and for the last 50 years at the University of California, Los Angeles (initially in the technical division and since 1970 in computer science) - to record the recent story of the area.

These periods largely coincide with what Professor Pearl called" the cause and effect revolution". "Three ascendant stages of what he refers to as the "leaders of causality" act as a key storyline for the " book of why. "The" revolution" in the book, in which Professor Pearl and his students play a decisive part, enabled scientists from a wide variety of fields to go beyond the first stage of the chain of causality in which they had repeatedly got bogged down.

"Depth neuronal nets have added many levels to the custom feature, but the rough information is still driving the adjustment process," says Professor Pearl. It was the cause and effect revolutionary that allowed scientists to study the higher levels of the ladders. Second stage of the causer's step is from seeing to doing.

Prof. Pearl observes that "many researchers have been traumatised to find out that none of the statistical techniques they have learnt is enough to raise a straightforward issue such as "What happens if we don't pay twofold? "The Book of Why" provides a comprehensive account and story of how and when a single cast alone can respond to such issues without the need for life-experiment.

One of the top rungs of the leaders contains so-called "counterfactual" questions: Formerly regarded as a formative feature of mankind, these issues have not escape the notice of ever more sophisticated modelling instruments in recent years. Professor Pearl, who emerged from the pages of "The Book of Why", is full of enthusiasm and proud of his fellow scholars and fellow Masters.

" Not all of the resulting book is fully available to those who do not agree with the author's preference for equation. In spite of this deliberate scepticism, Professor Pearl is remarkable in his optimism about what AI can do and even whether we can build machinery that can distinguish good from bad.

Irrespective of whether one is in agreement with these provoking inferences, we can all assume that in any contrafactual universe in which this is the case, the programmer with the charisma of Professor Pearl will have the say.

Mehr zum Thema