Professor Joseph Jacobson (MIT)

Artificial Intelligence and The Structure of the Written and Oral (Torah) Law

Artificial Intelligence is the framework for conferring capabilities, reminiscent of human-like intelligence, to computers. A sub-area of AI called machine learning (ML) is a mathematical framework for associating sets of things, such as a set of pictures of animals to a set of names of animals. In general there are two steps to machine learning. The first step, called training, consists of creating a set of correlated pairs of items from each set (a picture of an animal and its name, a picture of another animal and its name and so on) and running that through a computational procedure to generate a rule for correlating the sets of things (animals to their names). The second step called inference is a computational procedure for applying the rule to a new piece of data (a new picture of an animal) which predicts a corresponding label for that data (the name of that animal). Machine learning can also be used in the ‘reverse’, to, for instance, generate a picture of an animal when given the name of an animal. In this talk we plan to give a brief overview of how machine learning works along with some examples of what machine learning can do. Finally we draw a deep analogy between the framework of machine learning and the structure of The Written and Oral (Torah) Law.