Nonlinear causalities: the probem of (un)predictability in contemporary technology

Today we are witnessing an incredible technological development, especially in the field of artificial intelligence. This development gives rise to new philosophical reflections, questioning the old assumptions on understanding of machines, putting pressure on existent philosophies of technology. New theories are focusing mainly on explaining different effects of these technological novelties on human social and political life, from problematic psychological effects of new media to (anti)democratic effects in functioning of algorithms and many other changes. This project detects a gap in the philosophical study – not of the effects of technology, but rather of the functioning of its novel mechanisms. We propose a comprehensive study that would rethink the underlying philosophical concepts of the new mechanisms by confronting them with contemporary developments in the field of philosophy. This project aims to fill the theoretical gap with a study of fundamental philosophical categories which serves as a metaphysical base for the mechanisms of contemporary technology, with a focus on the idea of non-linear causality. The “old” philosophies of technology approached this phenomenon considering it as an automatized tool. But with recent developments in the field of artificial intelligence it became more complex: it is not only an automatized tool, it is also autonomous – modifying its own rules of functioning through machine learning. This calls for a new philosophical approach, one that analyzes these novel mechanisms. Machine learning and deep learning, big data analysis, predictive and prescriptive analysis – all of these are crucial mechanisms and they all rely on a metaphysical notion of non-linear causality. Any idea of non-linear causality implies a disturbance in a direct relation between cause and effect. The specific notion of non-linear causality demanded by the contemporary technology is called a feedback loop mechanism: a non-linear causality, where the cause A produces the effect B and because of a necessary divergence between the actual effect B and an intendend result of the cause A, the effect B, in its turn, produces an effect back on the cause A in order to adjust itself according to the intended result and eliminate or diminish the divergence. In this manner feedback loop mechanism attempts to control the unpredictability of events. However, this notion of feedback loop is part of a larger historical system of thought – the so called statistical reasoning. This reasoning replaced the idea of necessity with probability, thereby including the possibility of contingent events, which disturb the necessary relation between cause and effect and produce a variation in the effect. The mechanism of feedback loop presupposes the concepts of probability, it takes into account the variations and aims to control them through calculative procedures. But contemporary philosophy, focusing on metaphysical concepts of causality, got disconnected from theories of technology. Its development went in a different direction: instead of probabilities, contemporary philosophy attempted to outline the idea of radical contingency. Radical contingency is a specific notion of a non-linear causality, it doesn’t imply only a disturbance between cause and effect, but a complete disruption in the proces of causation itself – the effect itself is completely unpredictible. This project aims to analyze these new technological mechanisms, to place them in a historical theoretical context and to redefine their functioning; to confront core philosophical notions of new technological mechanisms with contemporary philosophy of contingency; and thus tackling the fundamental problem: if contemporary technology relies on the notion of probability and feedback loop mechanism which aims to control the unpredictability, what are its limits when we confront it with the concept of contingency which attempts to demonstrate the total unpredictability of events?