I am launching a new project in the area of Cognitive Computing Systems. Below is a brief description of the effort. You can read a recent post that I did for Brookings here. If you are interested in collaborating on this project, please contact me.

Project Description

Every major technology player is investing serious financial and human capital into the development of Cognitive Computing Systems (CCSs). CCSs leverage artificial intelligence and machine intelligence techniques to build systems that can 1) learn from interactions, 2) analyze large datasets in an effective and efficient manner, and 3) increase the of precision of outcomes overtime through continuous process of analyzing and learning from data. Put more simply, CCSs are inspired by the potential to mimic how the human brain acquires, analyzes, and employs data to make decisions. Research and development efforts and emerging prototypes point to the fact that CCSs have significant potential to cause major disruptions in all facets of the public sector. In addition, if one considers CCSs along with other emerging technologies (e.g. autonomous vehicles or fully connected machine-to-machine networks), the magnitude of disruption is potentially greater. Public agencies need to take a more proactive approach towards 1) understanding the nature of CCSs, 2) appreciating their potential to enable agencies to increase performance and optimize resource allocations, and 3) charting pathways towards adopting, experimenting with, and implementing these systems. This project will focus on providing public managers with actionable insights on the above-mentioned three issues.

Today, CCSs are being designed to support human decision-makers by providing them evidence-based solutions through analyzing large quantities of data within a domain. In an ideal world, according to technology enthusiasts, these systems should be able to make decisions on their own, thereby removing some of the traditional concerns with human decision-makers (e.g. prejudice, errors, etc.). CCSs will also be networked with other such systems so as to conduct transactions without human intervention. While there is no doubting that the sophistication of these systems will increase over time and their costs will decrease making them affordable, serious public policy and management considerations must be addressed to fully leverage their potential. Our project will focus on studying how public agencies can take a proactive approach to preparing themselves for CCSs. We will not only study the technical and organizational issues associated with CCSs, but take a close look at the social and public policy issues.

Our research methodology will take a multi-prong approach. First, we will document lessons learned from several of our ongoing projects that include elements of CSSs (e.g. machine learning algorithms, data mining, sensor-networks, etc.). Second, we will collect case studies on various CSSs projects in the public sector. We will seek to collect cases beyond the US, and will also look at CSSs projects at various scales. Through the analysis of these cases we will seek to uncover issues and challenges associated with implanting CCSs. In addition, these cases will enable us to tease apart various common themes that various stakeholders within the CCSs face as they conceive of, launch, and manage these projects. Third, we will leverage our vast network of public sector CIOs, technology enthusiasts, futurists, and innovators to conduct in-depth interviews on CCSs in the public sphere. Our interviews will focus on technical, organizational, policy, and social dimensions of CCSs and their role in 1) transforming public agencies, 2) innovating policy design, implantation, and evaluation, and 3) the nature of data, analytics, and systems as a public good.

Funding

The IBM Center for the Business of Government has generously provided support for the first report from this project.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.