We are excited to announce today that the John D. and Catherine T. MacArthur Foundation will be supporting a two-year initiative at Data & Society to better understand intelligent systems — automated systems that render judgment on data and exercise control in lieu of human management.
These technologies have proven increasingly ubiquitous in recent years, becoming a key part of the landscape and public dialogue in arenas as varied as capital markets, warfare, medicine, transportation, and social networks. Intelligent systems have the potential to disrupt many parts of modern life, prompting both dystopian and utopian rhetoric. In order to better understand these systems and their societal implications, it’s imperative to look more closely at the social, technical, and legal dynamics underpinning them.
The Intelligence & Autonomy (I&A) project will conceptualize these different systems as aspects of a more general phenomenon. The project will conduct research and convene forums in an effort to develop common frameworks and best practices for shaping the development of these technologies as a class in the public interest. Questions to be addressed by I&A include:
- To what extent do intelligent systems raise novel issues that require the recognition of new protections or rights within the context of the law?
- What should be the balance between technical standards and legal controls in the space? What aspects of intelligent systems are better managed by one or the other?
- What are generalizable design patterns that might guide technologists, engineers, and software developers as they create and implement these systems on a mass scale?
The project will be headed up by Data & Society fellow Tim Hwang. Hwang previously served as Chief Scientist of Pacific Social Architecting, a research collective exploring influence on social networks and bot behavior online. He recently completed a fellowship at the Stanford Center for Legal Informatics, focusing on automation and intelligent systems in the law. I&A builds on and expands a Microsoft-funded pilot project at Data & Society which developed the overarching approach of the project and comprises an ongoing comparative study examining existing intelligent systems regulatory practice.
In the first year, I&A will focus on three projects: 1) researching a series of case studies looking at the regulation of proto-intelligent systems historically, 2) creating a pattern library of technical best practices around the design of these systems, and 3) developing a set of simulation exercises to surface common points of policymaking tension across different arenas affected by intelligent systems. Based on lessons learned, the I&A project will turn toward applying these principles to the near-term economic and social challenges posed by these technologies.