A brilliant explanation of how and why AI needs to incorporate causation.
What is the difference between these two questions? First, does smoking cause lung cancer? Secondly, does the crowing of the rooster first thing in the morning cause the sun to rise? Well, we now know that the answer to question one is yes and the answer to question two is no. But we did not always know that. We probably guessed correctly the answer to the second and have done for years. But the first was a real conundrum. Believe it or not, for years nobody could prove that smoking caused lung cancer. Why – because the language of statistics did not acknowledge the existence of cause and effect. It only acknowledged correlation. Pearl’s book explains why knowing the answer to the question, why, is just as important as knowing the answer to the question, what.
Link to book: https://www.waterstones.com/book/the-book-of-why/judea-pearl/9780141982410
You may also like to browse other AI books: https://www.thesentientrobot.com/category/ai/ai-books/
