The Combinatorial Creativity of a Chat Bot. Tracing the Irreducibility of Human Actions Versus AI through Analysis of Defining Assumptions, Automatisms, Incapacities, and Computerized Evaluation Tests.

Are creativity and intelligence distinct, coincidental, or complementary concepts? Do they determine irreducible distinctions between human and artificial agent? Can they also imply automatism and inability as additional heuristic resources? In the distinctions between human and A.I., research on creative processes examines abilities, actions, behaviors, and conversations: elements that are disambiguable and measurable. However, it is possible to reverse the perspective, distinguishing human and artificial agent from inabilities, automatisms, and ambivalent behaviors. The irreducibility between human intelligence and A.I. depends on the starting theoretical and discursive assumptions accepted by researchers investigating these domains. Integrating ethnomethodology and discursive analysis, the present theoretical and empirical research highlights the irreducible differences between human intelligence, creativity, automatisms, and incapacities by analyzing theoretical assumptions and international tests used by programmers to evaluate chat bot interactions: a) The Classical Turing Test; b) The Inverse Turing Test; c) The Winograd Test; d) The Winogrande Text; and e) The Lovelace Test on the creativity of artificial agents.