Advanced NLP procedures as premises for the reconstruction of the idea of knowledge

This paper evaluates artificial texts produced by the GPT 2 and GPT 3 language models and proposes to use hermeneutic tools for their analysis, putting them on an equal footing with man-made texts. This decision is due to the intelligibility of these texts and the implicit knowledge on which they are based. Since the base of these language models is always the language corpus, it can be assumed that it is the source of this knowledge and that tools such as discourse and, in particular, the theory of discursive space can be used for its analysis.