Process-Oriented AI: Artificial Intelligence for a Connected World

Abstract

With the advancement of AI, using this technology to solve life challenges has become a real possibility. However, AI systems are laden with bias and often perpetuate old prejudices and injustices. To step beyond such limitations requires a leap outside the field to allow for a fresh perspective on its design. This dissertation looks at AI challenges through the lens of Alfred North Whitehead’s process philosophy to imagine AI design beyond its current state of development. The power of philosophy lies in its ability to bypass habitual perception and gain a fresh perspective.

This dissertation draws on process philosophy to develop a theoretical framework for a dynamic AI design that incorporates a multidisciplinary approach and more closely reflects human thinking. Key research questions addressed in this process are: What principles in Whitehead’s philosophy can be applied to AI? How can they be integrated into AI components, such as data and machine learning? Furthermore, what are the implications of designing AI systems through a process philosophy lens? How do these changes make a process-oriented AI different from traditional AI systems? How might this influence the relationship between humans and AI? Finally, how can complex technical ideas in this research be communicated without oversimplification or loss of detail? What are the possibilities of bridging communication gaps among disciplines?

The process-oriented AI system presented in this dissertation offers a dynamic design approach that influences and adapts to its environment. Rethinking AI design within the process philosophy framework involves taking gradual steps to combine empirical science with human stories, culture, and art. The outcome is a creative process involving human experts, who hold the domain knowledge, and a machine that is responsive to feedback and can efficiently interpret large amounts of information. This approach blends the quantitative analysis of what is happening with a more nuanced and complex view of why it is happening. The result is a co-creation, reconfiguration process that requires adjustment and collaboration, guided by human agents throughout its discovery.