Teleogenesis. A Bergsonian View on Problem-Finding in Human and Artificial Intelligence

Abstract

To understand human intelligence, problem-solving has frequently been conceptualised as a goal-oriented mental process, in which an agent transforms the ‘problem space’ to reach a goal-state: A hungry rat tries to obtain food and a chess player tries to win. From the standpoint of functionalism, goal-states like these are understood as image-like or propositional representations which are the subject of motivation. Due to the presupposition of a finalistic or mechanistic explanatory framework, what is rarely investigated is the origin of goal consciousness. From the point of view of such a framework, ‘all life is problem solving’ (Karl Popper), i.e., goal consciousness derives from a minimal logical discrepancy of two representations which is a universal type of relation. Yet, this reasoning falls short on processes of discovery which require forward search of goals. Human life is not exhausted by means-ends-analyses in a deductively closed field of operations. Differently put, human problem-solving entails teleogenesis. In it, human striving emerges, a process that may post hoc be described as teleological. Integrating this primordial constitution of goals into the understanding of problem-solving requires overcoming functionalist, mechanistic, or finalist presuppositions, which is only possible when revisiting the conceptual foundations. Bergson’s philosophy of creative evolution offers a critical perspective. According to his organistic standpoint, teleogenesis is different from a logical relation between two representations. Under the condition of new foundations, it might be possible to leave the conceptual impasse of problem- solving research in the human sciences, which has lasted for several decades.