DARPA program seeks to teach machines common senseNews
October 18, 2018
ARLINGTON, Va. The Defense Advanced Research Project Agency (DARPA) has launched the Machine Common Sense (MCS) program to explore recent advances in cognitive understanding, natural language processing, deep learning, and other areas of artificial intelligence (AI) research in an effort to crack the code of a critical component lacking in AI ? human-style common sense.
Under the new effort, the agency will focus on two approaches for developing and evaluating different machine common sense services. The first approach will create computational models that learn from experience and mimic the core domains of cognition as defined by developmental psychology; the areas include the domains of objects (intuitive physics), places (spatial navigation), and agents (intentional actors). Under this prong of the effort, researchers will seek to develop systems that think and learn as humans do in the very early stages of development, leveraging advances in the field of cognitive development to provide empirical and theoretical guidance.
The other approach will build a knowledge repository capable of answering natural language and image-based queries about common sense phenomena by reading from the web, with the resulting capability measured against the Allen Institute for Artificial Intelligence (AI2) Common Sense benchmark tests. DARPA predicts that researchers will use a combination of manual construction, information extraction, machine learning, crowdsourcing techniques, and other computational approaches to develop the knowledge repository.
“The absence of common sense prevents an intelligent system from understanding its world, communicating naturally with people, behaving reasonably in unforeseen situations, and learning from new experiences,” said Dave Gunning, a program manager in DARPA’s Information Innovation Office (I2O). “This absence is perhaps the most significant barrier between the narrowly focused AI applications we have today and the more general AI applications we would like to create in the future.
During the first few years of life, humans acquire the fundamental building blocks of intelligence and common sense,” Gunning added. “Developmental psychologists have founds ways to map these cognitive capabilities across the developmental stages of a human’s early life, providing researchers with a set of targets and a strategy to mimic for developing a new foundation for machine common sense.”