THE CULTURAL, ECONOMIC, AND INSTITUTIONAL DETERMINANTS OF AI INFRASTRUCTURES AND THEIR CONSEQUENCES IN GLOBAL CONTEXTS
Abstract
Research problem- How do cultural values and institutional priorities shape artificial intelligence infrastructures in comparative national and global contexts. What are the known and unknown effects of comparative AI contexts for security. Proposed methods- Field research in six countries – varying in terms of political economic systems and types of innovation systems for AI – will enable process tracing of the cultural values and the economic and institutional priorities that shape AI infrastructures. Next, the proposed methods synergistically integrate field research with cutting edge big data models (for collecting and analyzing media, business, security, and policy reports) and will lead to the creation of predictive models at the intersection of social science, natural language processing (NLP), and machine learning (ML). The models will correlate actors’ values with AI infrastructures on all issues of importance and will examine relationships and spillovers across national borders. Trained on six cases, the models are expected to generalize well to all countries with known AI policies. The methods designed in this project will additionally predict consequences of new issues-topics or the infrastructure of hypothetical countries with varying specifications. Basic research contribution- 1. Predictive models of AI infrastructure employing the notion of values from the ground up in comparative contexts taking into account a diversity of micro concerns from business, society, government, militaries, and media and the way that these translate into infrastructures. 2. Tight integration of engineering and bigdata techniques with social science. 3. Analytic coherence to the polysemic, multitudinous, and complex aspects of AI – through recognizing patterns within and across countries Anticipated outcomes of the research- 1. A comprehensive interdisciplinary literature review on cultural values and infrastructural studies. 2. Social science contributions (as above) to science, technology, and innovation studies, 3. Innovative bigdata models at the intersection of NLP and ML for predicting AI infrastructural values and characteristics within and across countries. 4. Country reports on each of the six cases in study, including spillover effects and security implications. 5. A coauthored book length project on the project theme. 6. Several journal articles on the novel methods employed and the empirical findings. 7. Training for undergraduate, graduate, doctoral and postdoctoral students. 8. Periodic workshops and conferences demonstrating new methods and findings from the study, and expanding the network of scholars beyond George Mason working on these issues. Impact on DoD capabilities or broader implications for national defense- The theoretical approach connects with strategic culture models for national security. Predictive models will analyze the correlations between actors’ values and skills, and national security needs – both knowns and the unknown unknowns generated from our computational models. The analysis will spell out security implications from parametric and scenario variations in AI specifications.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 07, 2023
- Source ID
- FA95502210171
Entities
People
- Jatinder Singh
Organizations
- Air Force Office of Scientific Research
- George Mason University
- United States Air Force