요약
AI is not creative, and the debate is a distractionspringer.com
본문
AI is not creative, and the debate is a distraction- Curmudgeon Corner
- Published: 23 March 2026
- (2026)
- Cite this article
AI & SOCIETY
Aims and scope
Submit manuscript
AI is not creative, and the debate is a distraction
Download PDF
The question of whether artificial intelligence is “creative” has become one of the most persistent—and least illuminating—discussions in contemporary AI discourse. Each incremental improvement in generative systems is treated as an invitation to reopen the matter as if creativity were a technical property waiting to be unlocked by scale, data, or architectural refinement. This framing is mistaken. AI is not uncreative because it has not yet crossed some experiential or ethical threshold; it is uncreative because it is not a subject at all. It is an instrument—designed, trained, and deployed in the service of external interests and neither chooses nor understands. The ongoing fascination with AI creativity does not clarify the nature of machines; it distracts from a more uncomfortable reality about modernity itself: that we increasingly confuse optimized performance with agency, and output with autonomy.
Recent debates often proceed as though the problem were one of insufficiency. AI lacks lived experience, ethical reasoning, embodiment, or social embeddedness. Address these gaps, the argument goes, and creative autonomy might follow. Yet, as I have argued in Creativity in the age of AI: the human condition and the limits of machine generation (2024), contemporary AI systems are not failed subjects awaiting completion; they are non-subjects by design. Their goals are assigned, their parameters fixed, and their scope of action tightly governed. They do not act for reasons they recognize as their own nor could they meaningfully refuse to act at all. Whatever novelty they generate is inseparable from the objectives imposed upon them.
Proposals to embed AI more deeply in physical and social environments, while technically sophisticated, miss the point entirely. Work on shared and socially embedded worlds, for instance, promises richer interaction and more adaptive behavior, as Pezzulo, Parr, and Friston describe in Shared worlds, shared minds: strategies to develop physically and socially embedded AI (2025). But embeddedness is routinely mistaken for autonomy. Interaction does not confer interiority, and social coupling does not produce self-authorship. A system can be situated, responsive, and context-sensitive while remaining fully instrumental, however convincingly it simulates participation.
What is often celebrated as AI creativity is better described as functional performance under conditions of maximal constraint: creativity without ownership, expression without interiority, novelty without responsibility. Generative systems recombine and optimize within a space defined elsewhere, according to values they do not originate and cannot revise. This is not a defect to be remedied; it is the condition of their usefulness. A system capable of refusing its assigned ends would no longer be an instrument. The fantasy of creative AI persists precisely because it allows us to enjoy the outputs without confronting the asymmetry of control that makes them possible.
The stakes of this confusion extend beyond machines. AI is frequently framed as a radical rupture—a new form of intelligence forcing us to rethink agency. In fact, it is better understood as a culmination. Modernity has long privileged instrumental reason: the organization of action around efficiency, prediction, and utility maximization. AI does not introduce this logic; it perfects it. Its non-subjectivity is not an aberration but an extreme case, revealing what agency looks like once autonomy has already been stripped away, as I have argued in Transhumanism as capitalist continuity: branded bodies in the age of platform sovereignty (2025).
The debate about AI creativity feels both urgent and oddly hollow. By asking whether machines might one day become creative subjects, we avoid asking how readily we have accepted technical and institutional systems that reward output while eroding authorship. Human creativity, too, is increasingly evaluated in terms of engagement metrics, platform visibility, and market alignment. Decision-making is guided by infrastructures designed elsewhere, according to logics few individuals meaningfully influence. AI does not threaten to dehumanize us by becoming human-like; it reflects a version of agency modern societies already tolerate.
Ethical discussions often focus on downstream harms: overreliance, deception, emotional attachment, and the psychological risks associated with interacting with systems that simulate understanding without possessing it. Chandra’s team explored these risks in their 2025 paper at the ACM Fairness, Accountability, and Transparency Conference, noting that conversational agents create the illusion of relational understanding without genuine subjectivity. These concerns are real, but they remain secondary. The deeper issue is that AI normalizes a model of action without responsibility or accountability, fluency without ownership. The absence of a subject is not experienced as a loss; it is reframed as convenience.
Calls for careful handling and responsible deployment, while well-intentioned, underestimate the structural nature of the problem. Cappuzzo, in the International journal of linguistics, literature and culture (2025), argues that careful management alone cannot produce autonomy where none is permitted to exist. As long as AI remains embedded within regimes of platform sovereignty and capitalist continuity, its lack of subjectivity is not a bug but a feature. The system cannot become a creative agent without ceasing to be governable in the ways that make it economically and politically valuable.
The persistence of the creativity debate reveals more about our conceptual habits than about AI itself. We continue to speak as though agency were a spectrum on which machines might gradually advance, rather than a condition that depends on the capacity to constitute one’s own ends. In doing so, we mistake obedience for autonomy and performance for authorship. AI appears creative only because we have lowered the bar for what creativity is allowed to mean.
AI is not creative, and this is not a temporary limitation awaiting technical resolution. It is the logical outcome of systems designed to act without subjects. The real question is why we find this so easy to celebrate, and what it suggests about the version of agency modernity now finds acceptable.
Data availability
No datasets were generated or analysed during the current study.
References
Cappuzzo B (2025) Artificial intelligence: a great opportunity to be carefully handled. Int J Linguist Lit Cult 12(1):59–64
Chandra M, Naik S, Ford D, Okoli E, De Choudhury M, Ershadi M et al (2025) From lived experience to insight: unpacking the psychological risks of using AI conversational agents. Proceedings of the ACM FAccT Conference.
Lockhart ENS (2024) Creativity in the age of AI: the human condition and the limits of machine generation. J Cult Cogn Sci 9:83–88. https://doi.org/10.1007/s41809-024-00158-2
Lockhart ENS (2025) Transhumanism as capitalist continuity: branded bodies in the age of platform sovereignty. Humans 5(3):21. https://doi.org/10.3390/humans5030021
Pezzulo G, Parr T, Friston KJ (2025) Shared worlds, shared minds: strategies to develop physically and socially embedded AI. EMBO Rep 26(17):4197–4202. https://doi.org/10.1038/s44319-025-00549-8
Curmudgeon Corner
Curmudgeon Corner is a short opinionated column on trends in technology, arts, science and society, commenting on issues of concern to the research community and wider society. Whilst the drive for super-human intelligence promotes potential benefits to wider society, it also raises deep concerns of existential risk, thereby highlighting the need for an ongoing conversation between technology and society. At the core of Curmudgeon concern is the question: What is it to be human in the age of the AI machine? -Editor.
Author information
Authors and Affiliations
AuthorsContributions
This is a sole-authored work, and the author reviewed and approved this version of the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Lockhart, E. AI is not creative, and the debate is a distraction. AI & Soc (2026). https://doi.org/10.1007/s00146-026-02942-w
Received: 31 January 2026
Accepted: 15 February 2026
Published: 23 March 2026
Version of record: 23 March 2026
DOI: https://doi.org/10.1007/s00146-026-02942-w
Share this article
Profiles
- Ezra N. S. Lockhart View author profile