Publication

ABSTRACT: 

Cyber-social learning systems (CSLS) are purposeful socio-technical systems that learn. They learn how to perform much better over time, as manifested in continuous progress toward, and ultimately in the achievement and maintenance of, extraordinary levels of fitness for purpose. Learning in this sense is more like learning how to play the piano than learning that a proposition is true or learning what function generated given data. CSLSs are systems that learn to, and that then do, perform at virtuoso levels of quality.

Many of today’s most critical systems — for defense, healthcare, education, community services, transportation, energy and environment, etc. — are archaic, vastly under-performing, and unsustainable. The challenge is to put them on a path to becoming CSLSs. An audacious goal is to transform them into CSLS that exhibit dramatic improvements in performance within at most a decade or two. The results would include great reductions in cost and environmental impact while vastly improving the security, health, wellbeing, and quality of life of billions of people the U.S. and around the globe.

The question of how to achieve the transformation of today’s systems into cyber-social learning systems of the future was the subject of a series of workshops sponsored by the Computing Community Consortium, advisory to the National Science Foundation and other policy makers. In this talk, I will introduce the concept of cyber-social learning systems; the need for new scientific, engineering, and design foundations to enable their development; and a path toward such foundations based on convergent research that integrates computing, complex systems studies, the social, behavioral, and economic sciences, and other disciplines, along with test and evaluation in diverse domains of practice, to realize the vision of a world of interconnected cyber-social learning systems at scale.

Click here for more information on the SERC Talk.