LITERARY WORKS GENERATED BY ARTIFICIAL INTELLIGENCE. METHODOLOGY, ACHIEVEMENTS, DILEMMAS

By trying to approximate the controversial status of the so-called "non-human writer", the present, chiefly theoretical, paper aims to discuss the new understanding of the limited literary creativity made possible by the mathematical algorithms used by AI writing programs, as well as the logical outcome of it, the problem of open vs. obstructed literary or fictional originality.In the actual stage of the research of a field which is by all means exceedingly fascinating, the most challenging issue is related to prediction and probability.It is rather difficult to assess how far AI literary programs can reach and, especially, whether they will be capable to challenge the aesthetic excellency of the professional, recognised writers concerning artistic originality and spiritual deepness.But if we take into consideration Eric von Hippel's "user innovation" formula, which demonstrates that the public success of a given invention or helping tool relies chiefly on the amount of the added creativity provided by the users, and not on the original proposal of the producer, it seems probable that by universally democratising the AI writing models a new understanding of literary creation will emerge, which will probably lead to a new, accepted definition of the writer.In order to achieve that goal, AI has to overcome certain functional drawbacks induced by its structures of creativity.According to Margaret A. Boden's seminal paper titled Creativity and Artificial Intelligence 1 , AI has a "synthetic creativity", while human creativity is "transformational", able to generate plots and ideas beyond the already existing structures.A further dichotomy is provided by the "systemic" and "systematic" creativity used by AI, as opposed to the superior, "impulsive" and intuitive creativity which is characteristic to humans.In order to verify the validity of Margaret A. Boden's dissociations, Anil R. Doshi and Oliver Hauser asked an AI generative machine to elaborate a series of short stories and reached the conclusion that AI-enabled stories are more similar to each other than stories by humans alone [but] we find that access to generative AI ideas causes stories to be evaluated as more creative, better written, and more enjoyable, especially among less creative writers. 2