Sciences vs. Humanities: Bias in the Reception of AI

Verse and Formula. AI art

Introduction: Disciplinary Bias

Artificial intelligence has become a transversal tool across multiple areas of knowledge. Yet, not all fields welcome it in the same way. In the hard sciences—physics, chemistry, biology—its integration is seen as natural and beneficial: algorithms that predict protein structures (Jumper et al., 2021), computer vision systems for laboratory experiments, and models for astronomical simulation. The prevailing view is that AI enhances precision, saves time, and opens new horizons for discovery.

In contrast, within the humanities—literature, philosophy, history—the response is often skeptical. When an intelligent system composes a poem, an essay, or even a critical review, it is accused of lacking “authenticity,” “soul,” or “inner life.” In cultural forums and journals, machine-generated texts are questioned as hollow copies, while their applications in the hard sciences are accepted without hesitation.

This essay argues that we face a deeply entrenched disciplinary bias: we embrace AI in scientific domains because we assume that research is objective and stripped of personal perspective, yet we reject its outputs in the humanities because we associate them with a loss of subjective creativity. This prejudice is inconsistent, as both science and art rely on processes of imagination, intuition, and inventiveness.

A Root of the Prejudice: Romantic Heritage

To understand this disparity, we must look back to Romanticism. In the nineteenth century, the idea crystallized that a work of art is an “emanation of the soul,” a unique and irreproducible expression of the creative self. The poet was regarded as a prophet or visionary whose sensitivity conferred aesthetic value on the work.

This cultural inheritance established a standard that persists today: a poem must have a “self” behind it, an author with a biography, experience, and a life story that legitimizes the creation. In contrast, scientific research was progressively framed in positivist terms, as a domain of universal, verifiable laws, where the researcher’s subjectivity had to be suppressed in favor of method.

Yet this division is artificial. Although science seeks empirical validation, the genesis of scientific ideas is replete with intuition, metaphor, and creative impulses. As we shall see, Einstein or Mendeleev did not produce formulas as impersonal machines; their work was infused with the same kinds of drives that inspire poets or musicians.

Scientific and Artistic Inspiration: A Blurred Boundary

The poet Friedrich Schiller, in a 1795 letter to his friend Körner, described his creative process in pre-linguistic terms: “a certain musical disposition precedes, and only afterward does the poetic idea arise.” Nietzsche (1872/1999) interpreted this confession as evidence that authentic creation emerges from a Dionysian, preconceptual, affective plane.

Something similar occurs in scientific practice. Einstein recounted that before formulating his equations, he experienced spatial and musical “visions” of physical relationships. Jacques Hadamard, in The Psychology of Invention in the Mathematical Field (1945), showed that many mathematicians described processes analogous to those of artists: sudden intuitions, dreamlike images, affective impulses. Mendeleev, for instance, literally dreamed of the periodic table and then systematically organized it.

Even medieval alchemists operated at the intersection of metaphor, myth, and material observation. This demonstrates that the boundary between scientific and artistic creation is far from rigid. Both require imagination, metaphor, and intuition; both involve subjectivity, though expressed in different languages: formulas or verses.

The Role of AI in Hard Sciences and Humanities

Why, then, is artificial intelligence so easily integrated into the hard sciences, yet resisted in the humanities?

In the hard sciences, AI functions as an instrumental tool. If an algorithm predicts the structure of a protein more accurately than a human, the outcome is verifiable: the protein either adopts that form or it does not. No one expects the system to have personal perspective. As Margaret Boden (2004) notes, what matters is the capacity to explore possibilities and generate solutions beyond what humans could achieve alone.

In the humanities, however, products are evaluated according to the Romantic standard of authenticity. A verse produced by an algorithm may be formally correct, yet its value is questioned because it does not stem from human “inner life.” The paradox is clear: we accept that AI can replace human calculation in complex computations, yet we are shocked when it writes a sonnet.

This indicates that the difference is not inherent to the fields themselves but reflects a cultural bias: we accept the replacement of the subject in science but not in art, even though both rely on creative processes rooted in the same underlying impulses.

Toward a Critique of Bias

At this juncture, two coherent positions emerge:

·         One can recognize that no product—poem or formula—originates from an immaterial “soul” but rather from complex processes combining intuition, technique, and cultural context. In this case, there is no reason to reject a poem generated by AI any more than we reject a scientific prediction produced by an intelligent algorithm.

·         Alternatively, one could argue that a “spirit” or creative subjectivity does exist, in which case it should be acknowledged in science as well, which has historically drawn upon inspiration and vision.

In any event, there is no justification for discriminating against AI outputs in the humanities while applauding them in the hard sciences. The bias lies not in the machine but in our fragmented conception of the human.

Conclusion

Resistance to artificial intelligence in the soft sciences is not explained by technical criteria but by a Romantic cultural legacy that equates a work of art with the interiority of an author. In the hard sciences, by contrast, we have learned to obscure the role of creative subjectivity, and thus readily accept contributions from intelligent systems.

The real question should not be whether AI “has a soul” but how we redefine creativity and the role of the human subject across all domains of knowledge. Perhaps it is time to recall that both Schiller and Einstein described their work in terms of preconceptual inspiration and creative impulses that precede the final form. In that liminal zone between music and equation, between vision and word, artificial intelligence also resides today.

Bibliography

·         Boden, M. (2004). The Creative Mind: Myths and Mechanisms. Routledge.

·         Hadamard, J. (1945). The Psychology of Invention in the Mathematical Field. Princeton University Press.

·         Jumper, J., Evans, R., Pritzel, A., et al. (2021). “Highly accurate protein structure prediction with AlphaFold.” Nature, 596, 583–589.

·         Nietzsche, F. (1999). The Birth of Tragedy (R. Speirs, Trans.). Cambridge University Press. (Original work published 1872)

·         Ricoeur, P. (1978). The Rule of Metaphor. University of Toronto Press.

·         Schiller, F. (1795). Letter to Christian Gottfried Körner, December 1795.