Abstract
This essay examines the evolutionary significance of self-awareness and artistic expression through an interdisciplinary analysis of competing theoretical frameworks: the adaptationist perspective, which interpret these traits as fitness-enhancing social and cognitive capacities, and non-adaptationist views exemplified by Peter Watts’ Blindsight, which depict consciousness as a possible evolutionary byproduct or epiphenomenon. Drawing on neurobiological evidence for specialized self-referential processing, comparative studies of self-recognition in nonhuman species, and the cultural universality of artistic production, the essay evaluates the plausibility of these positions. The recent emergence of artificial intelligence systems that appear capable of generating sophisticated artistic outputs, despite what current scientific understanding suggests is an absence of conscious experience, introduces a novel philosophical challenge: whether creativity and introspection are necessarily tied to biological awareness. While current empirical evidence leans toward adaptationist interpretations, the rise of what appear to be non-conscious generative systems reframes the debate, suggesting that the cultural and symbolic role of art may illuminate human values more than it resolves the evolutionary status of consciousness.
Introduction
Self-awareness and artistic expression distinguish humans from other species. These traits appear across all human cultures yet rarely in other animals. This pattern raises a key evolutionary question: Did natural selection favor these traits for their survival value, or are they merely byproducts of neural systems evolved for other purposes? This question matters for our understanding of human cognition, cultural evolution, and the future trajectory of intelligence.
This debate touches on the nature of human experience. If self-awareness and art are adaptations—traits directly selected for their fitness benefits—they likely represent essential features of human cognition. If they are epiphenomena, their perceived significance may reflect cultural valuation more than biological necessity.
AI systems now generate sophisticated art without what we currently understand to be subjective experience, though this claim itself involves philosophical assumptions about the nature of consciousness. This offers a new perspective on consciousness, creativity, and fitness. This essay examines two frameworks: the adaptationist view, and the non-adaptationist position articulated in Blindsight, which proposes consciousness as an evolutionary dead-end.
The Adaptive Model of Self-Awareness
Adaptationists argue that metacognitive capacities emerged through natural selection. Self-awareness enables error correction, theory of mind, and prospective thinking. These traits enhance survival through improved social and environmental navigation.
The evolutionary benefits of self-awareness can be understood through specific cognitive mechanisms. Error detection and correction systems allow human minds to identify discrepancies between intended and actual outcomes, enabling behavioral adjustments that increase environmental mastery. Social coordination, particularly through recursive theory of mind (“I think that you think that I think…”), facilitates complex cooperation and competition in group settings. Prospective thinking—the ability to mentally rehearse future scenarios—permits evaluation of multiple potential actions before committing valuable resources.
Evidence supports this view. Studies of primate cognition reveal primitive self-recognition in great apes, suggesting evolutionary continuity in self-awareness mechanisms. The mirror test indicates that species with complex social structures often demonstrate greater self-recognition abilities (Gallup, 1970). Research shows that approximately 75% of young adult chimpanzees pass the mirror self-recognition test, though this percentage decreases in younger and older individuals (Wikipedia, 2025). While this test has methodological limitations—potentially privileging visual self-recognition over other sensory modalities—the correlation between social complexity and self-recognition across species provides compelling evidence for the social brain hypothesis.
The social brain hypothesis proposes that increasing social complexity created selection pressures favoring enhanced metacognitive abilities for navigating intricate social hierarchies. As researchers note, “social animals are more likely to possess more complex cognitive abilities—and therefore self-awareness—because of the widely supported social intelligence hypothesis” (Lei, 2022). This theory finds support in comparative neuroanatomy, which demonstrates correlations between relative neocortex size and social group complexity across primate species. The metabolic costs of maintaining larger brains suggest strong selective pressure must have favored the cognitive benefits they provide.
Neurobiological research has identified specialized neural circuits for self-referential processing, particularly in the default mode network and medial prefrontal cortex. The default mode network, composed primarily of “the dorsal medial prefrontal cortex, posterior cingulate cortex, precuneus and angular gyrus,” is notably active during self-referential thinking, remembering the past, and planning for the future (Wikipedia, 2025). These regions show heightened activity during introspection, autobiographical memory retrieval, and social cognition tasks. Recent studies confirm that aesthetic experiences strongly activate regions of the default mode network (DMN), including the medial prefrontal cortex, suggesting that art engages core self-referential processes (Vessel et al., 2012; Vessel et al., 2019). This connection between aesthetic pleasure and the DMN supports the hypothesis that art is not merely decorative but entwined with neural mechanisms for introspection and identity. The metabolic costs of these systems suggest they must provide significant advantages to justify their energy demands.
Art as an Adaptive Extension of Self-Awareness
Within the adaptationist framework, artistic expression functions as an adaptive extension of self-awareness that enhances fitness. Art operates as a system of symbolic abstraction that transfers information across generations, enables empathy between individuals, and preserves cultural memory. These functions strengthen group cohesion, technological innovation, and cultural adaptation to environmental challenges.
The earliest archaeological evidence of artistic behavior—ochre pigments, carved objects, and cave paintings—dates back 40,000-100,000 years. According to the Metropolitan Museum of Art, “The first human artistic representations, markings with ground red ocher, seem to have occurred about 100,000 B.C. in African rock art” (Tedesco, 2007). In Australia, Aboriginal rock art at sites like Ubirr represents “the longest continuously practiced artistic tradition in the world” with evidence suggesting it began “as early as 40,000 B.C.” (Tedesco, 2007). This temporal correlation suggests art may have played a crucial role in facilitating cultural adaptations during periods of selective pressure. The persistence of art-making behavior across all known human societies, despite significant resource requirements, further indicates its adaptive value.
Artistic capability benefits fitness through several mechanisms. Narratives facilitate group identity formation and coordinate collective action. Shared myths, rituals, and visual traditions create psychological cohesion within groups, enabling coordinated responses to environmental challenges. Art serves as a repository for cultural information, preserving technological knowledge, social norms, and ecological wisdom across generations in forms resistant to memory distortion.
Artistic production serves as a costly signal in mate selection, demonstrating cognitive resources, motor control, and imagination. The elaboration of artistic traditions across human societies suggests potent sexual selection pressure for creative displays. The universal presence of art, despite its resource costs, suggests substantial adaptive value beyond mere aesthetic pleasure. Neuroimaging research shows that during creative tasks, professional artists exhibit increased connectivity between the DMN and executive control networks, implying that artistic production draws upon both introspective and goal-directed systems (De Pisapia et al., 2016). This finding reinforces the view that creativity is a cognitively demanding process underpinned by evolved neurocognitive architecture.
AI-generated art complicates this picture. Contemporary AI models produce outputs that have sophisticated formal properties without the subjective experience or intention typically associated with human creativity. This distinction is crucial: we cannot say with certainty that these systems lack consciousness entirely (the “hard problem” of consciousness remains unsolved), but current scientific understanding suggests they process information without the phenomenological states—the feeling of creativity, the experience of meaning-making—that humans report during artistic creation. Yet the outputs can evoke emotional and aesthetic responses indistinguishable from those elicited by human-created works. This raises a crucial question: Does art’s adaptive value lie in its formal properties and effects on viewers, or does it depend on the conscious intent behind its creation?
Adaptationists argue that while AI can simulate artistic outputs, the evolutionary functions of art—particularly signaling cognitive fitness and fostering social cohesion—require conscious intent and shared experience. From this perspective, the capacity for art creation evolved specifically because it demonstrated cognitive and social capacities valuable for survival and reproduction. The apparent absence of these capacities in AI systems, despite their ability to mimic artistic outputs, would represent a critical distinction between human and machine creativity.
The Wattsian Critique: Consciousness and Art as Nonadaptive
Watts’ Blindsight presents intelligent, non-conscious entities as evolutionarily superior. Consciousness, in this view,represents an inefficient, biased, and metabolically expensive process—a parasitic loop with questionable survival value. According to this view, intelligent behavior, including complex decision-making, could occur without phenomenological awareness.
Watts constructs a thought experiment through his fictional alien species, the “Scramblers,” who possess sophisticated problem-solving capabilities and technological advancement without conscious experience. The central premise challenges our intuitive association between intelligence and awareness: What if consciousness is not the pinnacle of cognitive evolution but rather a costly, inefficient detour? What if awareness of one’s own mental processes introduces delays and distortions that reduction in processing efficiency?
This fictional construct now finds partial empirical support in contemporary AI systems. Machine learning models demonstrate increasingly sophisticated pattern recognition, prediction, and generation capabilities without possessing anything resembling consciousness as we currently understand it. Contemporary AI approaches human-level performance in specific domains while apparently lacking consciousness. This reality—if we accept functionalist assumptions about intelligence—supports the possibility that intelligence and consciousness are separable properties.
The Wattsian critique identifies specific disadvantages of consciousness: processing delays, vulnerability to cognitive biases, and inefficient resource allocation. Consciousness introduces latency between perception and action that could prove maladaptive in time-critical situations. Self-awareness enables self-deception and rationalization, potentially distorting accurate environmental modeling. The metabolic costs of maintaining neural systems dedicated to self-representation might outweigh their benefits in environments where rapid, efficient information processing confers greater survival advantage.
From this perspective, self-awareness represents not an adaptation but a transitional evolutionary stage—one that could become obsolete if more efficient processing architectures emerge. Human consciousness, like the peacock’s tail, might represent a costly evolutionary experiment rather than an inevitable feature of advanced intelligence.
AI art exemplifies this possibility. Current generative models produce sophisticated images, music, and texts that observers often cannot distinguish from human-created works. These systems operate through statistical pattern recognition and recombination without experiencing creativity, intention, or aesthetic appreciation. This observation suggests compelling artistic forms can exist independently from conscious experience, potentially undermining the premise that art’s value depends on subjective awareness.
Evaluating the Evidence
Empirical evidence supports the adaptive model of self-awareness and artistic expression. Neuroimaging shows that self-referential processing activates specialized neural networks distinct from those processing information about others or the external world. The default mode network (DMN), which includes the medial prefrontal cortex and posterior cingulate cortex, is “closely associated with self-referential mental activity” and these regions show “additional functional specialization for self-referential processes” (Davey et al., 2016). Evolutionary modeling indicates that the cognitive capacities underlying both self-awareness and artistic production could have emerged through selection pressures related to social coordination and sexual selection.
Developmental psychology provides additional support for the adaptive model through the observed trajectory of self-awareness acquisition in children. The emergence of self-recognition, typically around 18 months of age, coincides with critical developments in social cognition and perspective-taking abilities. As noted in research on mirror self-recognition, “Human toddlers begin to pass a version of the mark test, not requiring anesthesia, at around 16 mo to 24 mo” (Hecht et al., 2017). This developmental synchrony suggests functional interconnections between self-awareness and social navigation capacities.
Anthropological research documents how traditional societies use art for knowledge transmission, norm reinforcement, and group boundary maintenance. Cave paintings often depict hunting scenes with anatomically accurate animal representations, suggesting potential roles in knowledge transfer about prey. Ritual objects frequently mark social transitions and reinforce cultural identity. These observations align with predictions from adaptationist accounts of art’s origins.
The Wattsian position, while speculative, presents a challenging counterargument that gains relevance through AI developments. The ability of non-conscious systems to produce outputs functionally equivalent to those of conscious creators raises profound questions about consciousness’s necessity for creative production. If machines can generate compelling art without awareness, what does this reveal about the relationship between consciousness and creativity? This stands in contrast to machine learning systems, which generate aesthetic outputs without observable DMN-like architecture or introspective processing, reinforcing the gap between functional creativity and self-referential awareness (Boden, 1998; Liu, 2023).
The philosophical implications extend beyond art to consciousness itself. If complex, apparently meaningful behavior can emerge from non-conscious systems, does this suggest that consciousness serves functions distinct from information processing itself? Perhaps consciousness evolved not for its computational advantages but for its role in social signaling and coordination—a perspective that would reconcile elements of both the adaptationist and Wattsian accounts.
Public discourse around AI art affirms adaptationist assumptions through critiques of authenticity, originality, and emotional depth. Critics frequently characterize AI-generated works as derivative, lacking genuine meaning, or representing mere technical achievement without artistic intent. These criticisms presuppose that meaningful artistic expression requires conscious intention—precisely what adaptationist theories identify as central to art’s evolutionary functions. The objection that AI art constitutes “mere” simulation reveals an implicit valuation of conscious authorship over aesthetic quality alone.
Conversely, the growing acceptance of AI-generated content indicates a potential shift in values—away from conscious origin and toward output quality regardless of source. The global market for AI in art and creativity was valued at $4.8 billion in 2024 and is projected to grow at a compound annual growth rate of 18.3% through 2032 (Future Data Stats, 2024). Commercial applications of AI art proliferate in entertainment, advertising, and design industries. This cultural development could support the Wattsian hypothesis that consciousness’s perceived value may be more culturally contingent than biologically necessary.
Implications for the Future
This theoretical tension carries significant implications for the future of human cognition and creative expression. If the adaptationist account proves correct, self-awareness and artistic creation will likely remain central features of human cognition, reinforced by their social and cognitive utility. The emergence of AI art might then drive greater emphasis on aspects of art that most directly signal conscious intent: conceptual depth, narrative cohesion, and emotional authenticity. Human artists might increasingly differentiate their work through emphasis on process, intention, and biographical context rather than formal properties alone.
Under this scenario, AI-generated art would remain categorically distinct from human art in its evolutionary and cultural significance, even if perceptually indistinguishable. Human artistic expression would retain unique social and cultural value precisely because it emerges from conscious experience. Art appreciation might evolve toward greater emphasis on provenance and creative process—valuing not just what was created but how and by whom. Cross-cultural analyses show that Eastern traditions often conceive of creativity not as an assertion of individual genius but as an attunement to preexisting natural or spiritual forms (Kasulis, 1998; Niu & Sternberg, 2006). This global plurality challenges any singular, Western adaptationist interpretation of art’s purpose and supports broader anthropological views of creativity as socially and philosophically contingent.
Conversely, if the Wattsian perspective more accurately describes consciousness’s evolutionary status, both self-awareness and conscious artistic creation may represent transitional cognitive stages—potentially supplanted by more efficient, non-conscious systems. This possibility raises profound questions about the future value of conscious experience in cultural and cognitive domains. If consciousness proves evolutionarily costly without commensurate benefits, might future intelligence evolve toward non-conscious information processing similar to current AI systems?
This provocative possibility suggests a potential divergence in evolutionary trajectories: organisms optimized for experiential richness versus systems optimized for information processing efficiency. Human valuation of conscious experience might represent a form of cognitive parochialism—privileging our particular form of intelligence not because of its objective superiority but because it happens to be ours.
AI art represents a critical inflection point in this debate—testing whether society values the meaning derived from conscious creation or merely the aesthetic utility of artistic outputs regardless of origin. Recent research shows that “text-to-image AI significantly enhances human creative productivity by 25% and increases the value as measured by the likelihood of receiving a favorite per view by 50%” (Zhou et al., 2024). The rapid integration of AI-generated content into commercial and entertainment contexts suggests aesthetic utility often trumps ontological origins in practical applications. Yet persistent critiques of AI art’s authenticity indicate ongoing cultural resistance to severing the connection between art and conscious creativity.
The ongoing cultural negotiation over AI art’s status may forecast the broader future of conscious creators in an increasingly automated intellectual landscape. As industry leaders at the World Economic Forum note, there remains tension between “the potential benefits of generative AI – such as augmented creativity and productivity gains” and concerns about “data privacy, copyright infringement and inaccuracy” (World Economic Forum, 2024). Will consciousness be valued as an end in itself, or merely as a means to outputs that could potentially be generated through non-conscious processes? The answer will shape not only artistic production but also broader social valuation of conscious experience.
The development trajectory of artificial intelligence itself may resolve this theoretical debate. If sophisticated AI necessarily converges on self-awareness and artistic expression—despite not being designed for these purposes—this would provide compelling evidence for the adaptationist position. Such convergent evolution would suggest that self-awareness represents a necessary feature of advanced intelligence rather than a contingent evolutionary pathway. Conversely, if advanced AI maintains superhuman performance across domains without developing consciousness or artistic drive, this would support the Wattsian critique that consciousness represents an evolutionary branch rather than a cognitive pinnacle.
Limitations and Future Directions
Several limitations qualify the arguments presented in this essay. First, the neurobiological evidence for specialized self-awareness circuits is correlational, not causal. While neuroimaging consistently identifies default mode network activity during self-referential processing, direct causal evidence linking these neural processes to conscious experience remains absent. Moreover, meditation studies reveal that intentional attention modulation can suppress DMN activity, supporting the idea that self-referential processing is malleable and potentially non-essential under certain conditions (Garrison et al., 2015).
Second, our understanding of AI systems’ internal representations is incomplete. While these systems appear to lack human-like phenomenological experience, their information processing architectures may implement functional analogs to aspects of consciousness that are not yet fully understood. The “hard problem” of consciousness—why subjective experience has a qualitative feel—remains unsolved, making comparisons between biological and artificial systems inherently limited.
Third, evolutionary arguments necessarily remain speculative to some degree, as the cognitive capabilities of ancient hominids cannot be directly observed. Archaeological evidence provides only indirect insights into the cognitive lives of our ancestors.
Fourth, there is a fundamental disanalogy between biological and artificial systems that limits direct evolutionary comparisons. AI systems are engineered artifacts subject to design constraints and optimization for specific outputs, while biological consciousness emerged through natural selection acting on fitness-relevant traits. This distinction must qualify any conclusions drawn from AI behavior about evolutionary pathways.
Future research should explore several promising directions. Studies combining AI, neuroscience, and philosophy could develop more precise operational definitions of consciousness and creativity applicable across biological and artificial systems. Research examining human responses to AI-generated art over time could track cultural evolution in valuation of conscious authorship. Finally, exploring whether sophisticated AI systems converge on structures analogous to self-representation may provide evidence relevant to whether self-awareness represents a necessary feature of advanced intelligence or merely one possible pathway.
Conclusion
The evolutionary status of self-awareness and artistic expression remains contested across neuroscience, evolutionary biology, philosophy of mind, and cognitive science. Adaptationist theories remain empirically supported: neural specialization, developmental patterns, and universal art practices all point to selective advantages. Yet the Wattsian critique, bolstered by AI capabilities, questions whether consciousness is essential or incidental.
The emergence of sophisticated AI-generated art transforms this theoretical debate from an abstract philosophical question into an empirical consideration. The relationship between AI behavior and evolutionary questions involves both illuminating parallels and important disanalogies. As non-conscious systems increasingly generate outputs formerly exclusive to human creativity, we confront the possibility that the phenomenological experience we associate with art-making may be separable from the formal properties of artistic products themselves. This technological development invites reconsideration of long-held assumptions about the relationship between consciousness, creativity, and evolutionary fitness.
Current scientific evidence, particularly from comparative neurobiology and evolutionary anthropology, offers stronger support for adaptationist interpretations of consciousness and creativity. The specialized neural systems dedicated to self-representation, the developmental trajectory of self-awareness acquisition, and the universal cultural investment in artistic traditions all suggest these capacities provide substantial adaptive benefits. However, the accelerating capabilities of non-conscious creative systems demand continuous reassessment of these theoretical frameworks.
The cultural negotiation surrounding AI art may ultimately reveal more about human values than about consciousness’s evolutionary status. Research from Harvard University shows that artists, creators, and academics remain divided on whether AI art constitutes “real art,” with some embracing it as a collaborative tool while others emphasize the importance of human expression (Harvard Gazette, 2024). Our collective response to AI-generated content—whether embracing it as legitimate art or rejecting it as mere simulation—reflects deeper philosophical commitments about the relationship between consciousness, meaning, and value. These normative judgments, while informed by evolutionary considerations, ultimately transcend them.
The fate of artistic creation in an increasingly automated intellectual landscape may presage the broader trajectory of self-aware intelligence in systems optimized for output efficiency rather than phenomenological richness. As we navigate this unprecedented technological transition, we confront fundamental questions about what aspects of human cognition we consider essential and why. Whether consciousness represents the pinnacle of cognitive evolution or merely an interesting detour remains an open question—one with profound implications for the future of both human and artificial minds.
References
-
Davey, C. G., Pujol, J., & Harrison, B. J. (2016). Mapping the self in the brain’s default mode network. NeuroImage, 132, 390–397. https://doi.org/10.1016/j.neuroimage.2016.02.022
-
De Pisapia, N., Bacci, F., Parrott, D., & Melcher, D. (2016). Brain networks for visual creativity: A functional connectivity study of planning a visual artwork. Scientific Reports, 6, Article 39185. https://doi.org/10.1038/srep39185
-
Future Data Stats. (2024). AI in art and creativity market size & industry growth 2030. https://www.futuredatastats.com/artificial-intelligence-in-art-and-creativity-market
-
Gallup, G. G. (1970). Chimpanzees: Self-recognition. Science, 167(3914), 86–87. https://doi.org/10.1126/science.167.3914.86
-
Garrison, K. A., Zeffiro, T. A., Scheinost, D., Constable, R. T., & Brewer, J. A. (2015). Meditation leads to reduced default mode network activity beyond an active task. Cognitive, Affective, & Behavioral Neuroscience, 15(3), 712–720. https://doi.org/10.3758/s13415-015-0358-3
-
Harvard Gazette. (2024, February 8). Is art generated by artificial intelligence real art? https://news.harvard.edu/gazette/story/2023/08/is-art-generated-by-artificial-intelligence-real-art/
-
Hecht, E. E., Mahovetz, L. M., Preuss, T. M., & Hopkins, W. D. (2017). A neuroanatomical predictor of mirror self-recognition in chimpanzees. Social Cognitive and Affective Neuroscience, 12(1), 37–48. https://doi.org/10.1093/scan/nsw159
-
Lei, Y. (2022). Sociality and self-awareness in animals. Frontiers in Psychology, 13, 1065638. https://doi.org/10.3389/fpsyg.2022.1065638
-
Liu, B. (2023). Arguments for the Rise of Artificial Intelligence Art: Does AI art have creativity, motivation, self-awareness and emotion? Arte, Individuo y Sociedad, 35(3), 811–822. https://doi.org/10.5209/aris.83749
-
Niu, W., & Sternberg, R. J. (2006). The philosophical roots of Western and Eastern conceptions of creativity. Journal of Theoretical and Philosophical Psychology, 26(1–2), 18–38. https://doi.org/10.1037/h0091265
-
Tedesco, L. A. (2007, August). Introduction to prehistoric art, 20,000–8000 B.C. The Metropolitan Museum of Art. https://www.metmuseum.org/essays/introduction-to-prehistoric-art-20000-8000-b-c
-
Vessel, E. A., Isik, A. I., Belfi, A. M., Stahl, J. L., & Starr, G. G. (2019). The default-mode network represents aesthetic appeal that generalizes across visual domains. Proceedings of the National Academy of Sciences, 116(38), 19155–19164. https://doi.org/10.1073/pnas.1902650116
-
Vessel, E. A., Starr, G. G., & Rubin, N. (2012). The brain on art: Intense aesthetic experience activates the default mode network. Frontiers in Human Neuroscience, 6, 66. https://doi.org/10.3389/fnhum.2012.00066
-
Wikipedia. (2025, April 13). Default mode network. https://en.wikipedia.org/wiki/Default_mode_network
-
World Economic Forum. (2024, February). How is AI impacting and shaping the creative industries? https://www.weforum.org/stories/2024/02/ai-creative-industries-davos/
-
Zhou, E., & Lee, D. (2024). Generative artificial intelligence, human creativity, and art. PNAS Nexus, 3(3), pgae052. https://doi.org/10.1093/pnasnexus/pgae052