Are the Futures Computable? Knightian Uncertainty & Artificial Intelligence

Abstract

The growing sophistication of artificial intelligence (AI) tools promises to revolutionize organizing processes in the near future, transforming how firms identify, gather, analyze, and utilize information from their internal and external operating environments. The business venturing process, however, presents a unique set of challenges for AI tools since entrepreneurs often must navigate informational environments characterized by Knightian Uncertainty, defined as an unbounded set of future possibilities which cannot be measured or predicted. Because these uncertainties are by definition, unmeasurable using probabilistic tools, the presence of unmeasurable sources of Knightian Uncertainty logically raises questions about the long-term viability and relevance of AI tools to the field of entrepreneurship. In this study, we examine these questions through the lens of computational theory – namely the Church-Turing computability thesis – to examine the extent to which particular classes of AI tools are capable of addressing the unique epistemic problems intrinsic to Knightian Uncertainty in entrepreneurship. In doing so, our comprehensive analysis of the organizational implications of AI tools through the lens of modern computational theory allows us to identify the effective limits and boundary conditions of AI tools to navigate entrepreneurial environments characterized by Knightian Uncertainty. We conclude the paper with a robust future research agenda at the intersection of artificial intelligence and entrepreneurship theory.