DARPA Embraces ‘Common Sense’ Approach to AI

(Zhitkov Boris/Shutterstock)

The Pentagon’s top research agency is focusing its considerable AI efforts on the interim stage of machine intelligence between “narrow” and “general” AI.

The Defense Advanced Research Projects Agency (DARPA), which announced a multi-year $2 billion “AI Next” “campaign last month, is tightening its focus on teaching machine “common sense” reasoning. That capability remains beyond the reach of current AI constructs, and the research agency said it hopes to launch a “third wave” of AI technology that is adaptable while shedding light on the mystery of how machines learn.

Common sense reasoning is defined as “the basic ability to perceive, understand, and judge things that are shared by nearly all people and can be reasonably expected of nearly all people without need for debate.”

AI experts note the gap between AI inference and the ability to design systems that can draw directly on the rules of inference to achieve common sense reasoning. “Articulating and encoding this obscure-but-pervasive capability is no easy feat,” DARPA program managers note.

The lack of machine common sense is also seen as among the biggest barriers to advancing beyond narrow to general AI applications. The DARPA effort, dubbed “Machine Common Sense,” seeks to move beyond current frameworks that the agency considers “brittle” and lacking in semantic understanding.

Making perfect sense, DARPA’s Information Innovation Office kicked of its Machine Common Sense effort last week with a “Proposers’ Day” designed to brief potential bidders on program requirements. The key areas of focus are two-fold: building machines that learn from experience “like a child” and learning from reading like a “research librarian.”

The agency said it would fund research exploring developmental psychology, then establish a set of cognitive development milestones for determining how the resulting computational models learn in three areas: experience learning, prediction and “expectation” as well as problem solving.

A parallel effort will use web browsing to assemble a repository of machine common sense capable of answering queries based on natural language and images. The results will be tested against the Allen Institute for Artificial Intelligence benchmark. (Paul Allen, the Microsoft co-founder who died earlier this month, doubled the institute’s budget earlier this year to expand its research into machine common sense.

“Developmental psychologists have found 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,” said DARPA’s Dave Gunning.

These and other agency efforts are part of its more than $2 billion portfolio of AI research programs aimed at advancing beyond current technologies that require large amounts of training data, failed to adapt to change and “are unable to provide users with explanations of their results.” That last flaw has stimulated a market for research into “explainable AI” that seeks move away from the black-box style of predictive computing.

“Today, machines lack contextual reasoning capabilities, and their training must cover every eventuality, which is not only costly, but ultimately impossible,” noted DARPA Director Steven Walker. “We want to explore how machines can acquire human-like communication and reasoning capabilities, with the ability to recognize new situations and environments and adapt to them.”

Recent items:

Opening Up Black Boxes with Explainable AI

Sorting AI Hype from Reality