Representational Aspects of Artificial Intelligence Generated Images: Semiosis, Self-Referentiality, and Meta-Synthesis
This article aims to explore the relationships between the theoretical and conceptual operators of semiosis and self-referentiality, investigating aspects related to the representational dynamics of images generated by artificial intelligence (AI). The work begins with the ambivalent relationship be...
Guardado en:
| Publicado en: | URI:https://journals.openedition.org/cs, |
|---|---|
| Autores principales: | , |
| Formato: | Article ou chapitre numérique |
| Lenguaje: | Anglais |
| Publicado: |
Comunicação e sociedade
2025
|
| Materias: | |
| Acceso en línea: | Accès Université d'Orléans et IFPM |
| Sumario: | This article aims to explore the relationships between the theoretical and conceptual operators of semiosis and self-referentiality, investigating aspects related to the representational dynamics of images generated by artificial intelligence (AI). The work begins with the ambivalent relationship between the image as a criterion of truth and the recognition of its possibilities for manipulation to problematise how emerging aspects of contemporary media logic strain the supposed objectivity of technical images. In this sense, through its own conceptual vocabulary, it aims to characterise synthetic images (generated by AI) as metastases. This proposal is based on an analysis of the representational processes involved in the creation and circulation of AIgenerated images, examining the dimensions of prediction, self-referentiality, and meta-synthesis. Thus, the article discusses the specificities of representation and records of reality in technical images, deals with AI-generated images through prediction and their representational status, and characterises them based on the notions of “self-referentiality” and “meta-synthesis” in semiotic networks. In conclusion, the work considers that the imagistic metastases of AI-generated images strain the logical place of representation of the real, which is thus distanced, extended, and traversed by radical mediations and profound mediatisations. Likewise, AI-generated images create self-referential semiosis plots that are statistically predictable and probable, like other previous technical images. |
|---|