Interview with Pedro Domingos, the inventor of Markov Logic Network

In 2015, "The Master Algorithm" was published, providing an insightful and comprehensive look into the world of artificial intelligence (AI) and machine learning, particularly for those outside the field. Written by AI pioneer Pedro Domingos, the book showcases his expertise, having been involved in the field since the early 1990s. With numerous awards under his belt, such as the prestigious SIGKDD award and the invention of Markov Logic Networks, Domingos is well-equipped to discuss the evolution of AI, machine learning, and more.

In a recent interview with AIM, Domingos spoke about the development of AI over the years, from its humble beginnings to its current status as an indispensable part of many industries. He also discussed the challenges facing the field, including the difficulty of keeping up with its rapid expansion and the need to combat inertia.

"The Master Algorithm" was inspired by Domingos' desire to introduce the world of AI and machine learning to a broader audience. In the book, he presents the quest for the ultimate learning algorithm, which has the ability to learn anything. He also introduces the "tribes" of machine learning – Symbolists, Connectionists, Evolutionaries, Bayesians, and Analogies – each with its own approach to AI. Domingos firmly believes that a unified theory of machine learning can only be achieved by combining ideas from each of these paradigms.

When Domingos was first introduced to AI, it was through a book that sparked his curiosity. He recognized the potential of machine learning to revolutionize the world, but at the time, few colleges offered formal degrees in the field. He obtained his Ph.D. at the University of California, Irvine, which was one of the few institutions with substantial machine learning courses. Domingos considers Geoff Hinton and Tom Mitchell among his earliest heroes in the field.