By Lázara Ávila Fernández

Could it be that the concepts of author, authorship, and creation have become obsolete and need to be updated in light of these times, when language models occupy an ever-greater space in our lives? Should we authors refrain from using AI in our creative and literary production processes? Why is there resistance to acknowledging an increasingly widespread need for its use?

To address these and other questions, even if only indirectly, let us begin with what truly matters: the incompleteness of the sign, an inherent feature of the output that makes the Interpretant a prominent figure. Understanding its nature is neither a moral nor a puritanical obligation. It reveals that, despite its sophistication and speeds almost unimaginable to human beings, it does not reach the human cognitive horizon. Here, the cognitive is understood as that which pertains to or is related to knowledge and the mental processes that facilitate it, in their social, historical, and transcendental dimensions.

The Quasi-Sign

The language model always returns an incomplete sign to us: a quasi-sign that demands an Interpretant. As we explained when developing the Demanding False Zero (DFZ), this demand does not attribute consciousness or agency to the output, but rather points to its conceptual incompleteness. Recall that Charles Sanders Peirce’s semiotic triad constitutes a model that defines the sign as an interconnected relationship among three indispensable elements. These elements are the representamen (or sign), the Object, and the Interpretant. The representamen is the physical or mental object—a sound, a word, an image, or a smell—that stands in place of something else. The Object is that to which the representamen refers or alludes, whether a reality, an element, or an idea. The Interpretant, for its part, is not the person who interprets, but the new sign or effect—the idea or meaning—that is created in the mind of the recipient as a consequence of the representamen. The output, then, would be a quasi-sign because, in its production within the black box, the language model lacks a relation of its own to the Dynamic Object and to Secondness.

In this essay, as in our previous work, we do not align ourselves with the idea that the machine functions as a co-author, nor with the tendency that holds that machines possess cognition without consciousness. That would be to attribute to it a value we do not share, since we understand that there is no mind inside the language model, much less a process of semiotic comprehension.

Auctor: The One Who Makes the Fruits Grow

We bring in this word, rescued from Latin, because we are interested in answering the following: What are we facing when we speak of language models? Are we truly being displaced?

The term auctor, now in disuse, in its most ancient sense meant “the one who increases,” “the one who makes grow.” It also meant “the one who augments” and “the one who founds.”

In official Latin usage, auctor is also the one described as “qui in senatu primus sententiam dicit,” that is, the one who was the first to express an opinion in the Senate. Today, it could very well be the one who occupies the privileged place of sense, of the word grounded in the act of signifying. This, though redundant, is no less pressing.

And in law, this term, auctor, designates the guarantor. It is, then, the one who answers and acts responsibly. It derives from the verb augēre, which means to make grow, to increase.

But the matter is far more complicated than a simple enunciation or an aspiration conceived as an ideal. It means changing our understanding of what many take for granted: co-creation, the death of the author.

Note: This text is an excerpt from Augere: The Death of the Author. The English version was translated from the original Spanish with the assistance of AI-based translation tools and subsequently revised to preserve the book’s conceptual terminology, argumentative structure, and authorial intent.