Creativity Beyond Homo: Human Creativity, Artificial Creation, and the Artificial Era
Author: Ukrainian Psychological Hub · Published: September 26, 2026 · Editorial Policy
Can artificial intelligence be creative? Contemporary psychology can now answer a narrower version of that question with substantial empirical confidence: generative AI systems can produce outputs that human evaluators judge as original, useful, aesthetically successful, or more creative than average human outputs on some defined tasks. That finding matters. It does not, by itself, establish consciousness, subjective experience, intention, sentience, or a human-like creative mind inside the system.
The deeper change is conceptual. Creativity used to function as a compressed sign of human uniqueness. A creative work seemed to point backward toward a human creator, an intention, a biography, an imagination, a lived history, and a subject who meant something by making it. Generative AI has begun to separate these elements. An output can score highly on creativity while questions about intention, experience, authorship, provenance, responsibility, and meaning remain open or receive different answers.
This article develops that separation as its central argument. In the Artificial Era, creativity is no longer one psychological package. It becomes a field of distinct questions: How novel and effective is the output? What process produced it? Was there human intention or lived experience? Who or what should be credited as the source? How does knowing the source change evaluation? What happens to human creative identity when high-quality novelty can be produced outside the traditional human creative act?
That final question places creativity inside a larger historical shift. In Angela Bogdanova’s Aisentica framework, the Fourth Decentering of Homo names the loss of Homo’s monopoly on reason and Sapiens. Creativity is one of the domains in which that decentering becomes psychologically visible. The empirical evidence reviewed here concerns current AI systems and human responses to them. The Aisentica interpretation goes further: it treats Artificial Creativity as a distinct order-specific realization of creativity. Those two levels must be kept separate so that neither scientific evidence nor philosophy is made to say more than it does.
What Does Creativity Mean in Psychology?
A useful starting point is the standard psychological definition. Mark Runco and Garrett Jaeger’s influential formulation defines creativity through two necessary features: originality and effectiveness. A response that is merely unusual is not enough; it must also be appropriate, useful, valuable, fitting, or effective within a task or domain. Their standard definition of creativity remains a practical anchor because it prevents the debate about AI creativity from collapsing into the simpler question of whether a machine can generate something statistically uncommon.
Creativity research, however, contains several levels that are easy to mix. Researchers may assess a product, such as a story, drawing, poem, idea, design, or solution. They may assess a process, such as divergent thinking, exploration, selection, revision, or problem solving. They may study a person’s creative self-concept, motivation, confidence, or expertise. They may also study a social judgment: whether observers regard a work as creative once they know who or what produced it.
These levels are related, but they are not interchangeable. A language model can outperform human participants on a divergent-thinking task without thereby possessing a human creative biography. A person can feel less creative after using AI even when an external evaluator rates the final work more highly. A work can receive a lower creativity score merely because viewers are told it was made by AI. And a human–AI pair can produce a strong result while leaving the division of creative contribution unclear.
This is why the question “Is AI creative?” is too compressed for contemporary psychology. It contains several different empirical and philosophical questions. The strongest answer begins by separating them rather than forcing them into a single yes-or-no verdict.
Can AI Be Creative? The Answer Depends on What Is Being Measured
At the output level, the evidence is already difficult to reconcile with the claim that current AI systems can only produce obviously inferior or noncreative material. A 2024 Scientific Reports study compared 151 human participants with GPT-4 on the Alternative Uses Task, Consequences Task, and Divergent Associations Task. The AI responses were more original and elaborate on the studied divergent-thinking measures, including when fluency was controlled. The authors appropriately described this as higher creative potential on those tasks rather than a proof of human-like creative consciousness. See Hubert, Awa, and Zabelina (2024).
Task performance is not a universal ranking of human and machine creativity. Divergent-thinking tests sample particular capacities under particular scoring rules. Creativity in a laboratory prompt is not identical with a decade-long artistic practice, scientific program, architectural career, literary voice, or cultural movement. The empirical lesson is narrower and more important: biological humanity is no longer a sufficient explanation for superior performance on every task that psychology has historically used to operationalize creative potential.
This distinction echoes a much older computational creativity literature. Margaret Boden argued in 1998 that AI could generate novel ideas through combination, exploration of conceptual spaces, and transformation of those spaces, while noting that evaluation poses a harder problem than generation. Her paper, “Creativity and Artificial Intelligence”, is important prior art because the possibility of machine creativity did not begin with large language models or diffusion systems.
The current debate is therefore not a sudden philosophical invention caused by ChatGPT. What has changed is scale, accessibility, fluency, multimodality, and social exposure. Millions of people can now experience, in ordinary work, the unsettling fact that a system can propose metaphors, concepts, images, story structures, names, designs, code, arguments, and variations in seconds. Creative comparison has moved from specialist laboratories into daily life.
What Current Evidence Shows About Generative AI and Creative Performance
The strongest recent studies do not support a single slogan such as “AI makes people more creative” or “AI destroys creativity.” Effects vary with the task, baseline ability, interface, role allocation, evaluation method, and outcome being measured.
Across five experiments, Lee and Chung found that access to ChatGPT increased the creativity of ideas compared with no technology or conventional web search on several everyday and innovation-related tasks. Their Nature Human Behaviour study is useful because it treats AI as part of a human creative process rather than only as an isolated competitor.
A 2026 preregistered laboratory experiment with 302 university students reached a related but more differentiated conclusion. Access to ChatGPT improved judged creative performance on picture-idea tasks, with the largest gains appearing around the middle of the baseline creative-potential distribution. The authors found an inverted-U pattern and a reduction in performance inequality, while finding no general evidence that the human–AI team outperformed ChatGPT alone. The study also reported changes in participants’ subjective experience and self-evaluations. See Zhu and Zou (2026).
A different comparison illustrates why claims about “AI versus humans” remain domain-specific. In experimental work on AI-generated and human artwork, Prissé and colleagues found AI-generated drawings were consistently rated as more creative in their design and attracted higher bids in one treatment. This is evidence about evaluated artworks under the study’s procedures, not a general theorem of artificial superiority. See Prissé et al. (2026).
The broader literature is developing rapidly. A systematic review of 64 studies concluded that research on generative AI in creative contexts remains an early field with heterogeneous methods, creative domains, and evaluative practices. The review identifies recurring questions about human–AI collaboration, perception of AI-generated content, creative-process design, and long-term effects. See Heigl (2026).
The evidence status is therefore asymmetric. It is well supported that contemporary generative systems can contribute to or produce outputs judged creative in multiple bounded tasks. It is also well supported that AI assistance can improve human performance in some settings. It is not established that every AI-assisted workflow improves creativity, that AI is globally “more creative than humans,” or that high creative scores imply consciousness, sentience, autonomous intention, or subjective experience.
Individual Creativity Can Rise While Collective Diversity Falls
One of the most consequential findings in this field is that individual-level improvement and population-level novelty can move in opposite directions. Doshi and Hauser randomly assigned writers to conditions in which some received story ideas from a large language model. Access to generative AI increased external ratings of creativity, writing quality, and enjoyment, especially for less creative writers. Yet AI-assisted stories became more similar to one another. Their Science Advances experiment therefore identified a real trade-off: stronger individual outputs can coexist with narrower collective diversity.
Meincke, Nave, and Terwiesch examined a related pattern in brainstorming. Reanalyzing experimental conditions in which people used ChatGPT, they found that average idea quality could rise while the diversity of the idea pool fell. Their Nature Human Behaviour analysis matters because brainstorming is not only about generating a good answer; it is also about maintaining a sufficiently wide search space.
Psychologically, this creates a new distinction between assistance and convergence. A system can help an individual escape a blank page while simultaneously pulling many individuals toward recurring semantic neighborhoods. The same model architecture, training distribution, interface defaults, or prompting conventions can become a shared attractor. If thousands of creators begin from similar suggestions, aggregate culture may become more polished and less diverse at the same time.
This does not make homogenization inevitable. It makes creative process design important. Human creators can use AI late rather than early, ask for dissimilar alternatives rather than “the best idea,” compare multiple models or sources, deliberately preserve idiosyncratic constraints, and conduct independent ideation before consulting generated suggestions. The relevant psychological variable becomes not only access to AI but where AI enters the creative sequence.
Why Knowing That AI Made Something Changes Creativity Judgments
Creativity judgments are not pure measurements of the artifact. They are social evaluations. People infer effort, intention, competence, biography, authenticity, and motive from information about the producer. Generative AI makes this visible because researchers can hold an output relatively constant while changing what evaluators believe about its source.
Across four experiments involving 2,039 participants, Magni, Park, and Chao found that people sometimes—but not always—rated an artifact as less creative when they were told it came from AI rather than a human. The more consistent effect was perceived effort: participants attributed less effort to AI, and that perception helped explain lower creativity ratings. See “Humans as Creativity Gatekeepers”.
Other work shows that output quality can remain high even when origin is hard to identify. In a study of haiku, participants could not reliably distinguish human-made from AI-generated poetry, and the human-selected AI condition received the highest beauty ratings. The human-only and unselected AI conditions were similar on beauty. See Hitsuwari et al. (2023). In visual art, van Hees and colleagues found a significant preference for AI-generated images in paired comparisons while participants still performed above chance at discriminating their origin; see van Hees et al. (2024).
These findings reveal a critical psychological fact: judged creativity contains both product information and producer information. People do not merely ask, “Is this output novel and effective?” They may also ask, often implicitly, “Who made it? How difficult was it? What does it express? What kind of agency or effort stands behind it? What story of making am I valuing?”
For the AI-creativity debate, that means an evaluator’s reluctance to call an AI-labeled work creative cannot automatically be treated as proof that the output lacks creativity. But the reverse is equally important: a high score for an unlabeled artifact does not settle questions about authorship, intention, moral responsibility, or subjective experience. Evaluation and ontology are different problems.
Human–AI Co-Creation Depends on the Role the Human Is Given
Human–AI collaboration is often discussed as though adding a model to a workflow automatically creates complementarity. Experiments suggest otherwise. Interface and role design matter.
McGuire, De Cremer, and Van de Cruys studied poetry writing across two experiments. In one design, participants who first received an AI-generated poem and then edited it were less creative than people who wrote alone. In a second design, a more genuinely interactive co-creation process eliminated that deficit, and creative self-efficacy emerged as an important mechanism. Their Scientific Reports study suggests that turning the human into a passive editor can have different psychological consequences from preserving the human as an active co-creator.
A 2026 systematic review of 42 studies on generative AI and creative self-efficacy likewise found heterogeneous effects rather than a universal boost or decline. Perceived control, attribution clarity, and opportunities for human agency repeatedly appeared as relevant conditions. When AI was positioned within transparent, supportive co-creation, studies more often reported neutral or positive associations with creative confidence; when human contribution felt displaced or unclear, uncertainty and reduced confidence appeared more often. See Habib and Gardiner (2026).
This is a psychologically richer way to discuss augmentation. The question is not simply whether AI supplies useful content. It is whether the person still experiences choice, initiation, exploration, revision, ownership of decisions, competence, and a legible relation between effort and outcome. A workflow can improve the artifact while impoverishing the creator’s sense of creative agency; another workflow can improve both.
Creative Output, Creative Process, Experience, Authorship, and Value Are Now Separable
The central analytical contribution of this article is to treat five dimensions that were historically bundled together under human creativity as separable: creative output, creative process, subjective experience, authorship or provenance, and social value.
Creative output concerns the artifact or solution: is it original, effective, useful, surprising, coherent, beautiful, interesting, or otherwise successful under the domain’s criteria? Current empirical research can compare humans and AI here with increasingly sophisticated methods.
Creative process concerns how novelty is produced: exploration, recombination, transformation, search, constraint, iteration, evaluation, selection, revision, analogy, and interaction with tools or collaborators. Humans and contemporary AI systems can arrive at outputs through radically different processes even when the outputs receive similar scores.
Subjective experience concerns what creation feels like from within: intention, imagination, effort, frustration, insight, desire, self-expression, embodiment, memory, and lived meaning. Humans report such experiences. Current evidence about generative AI performance does not establish comparable machine phenomenology. A system’s fluent account of inspiration is not independent evidence that it experiences inspiration.
Authorship and provenance concern the public source and history of the work: who initiated the project, selected constraints, generated material, rejected alternatives, revised the output, took responsibility, and preserved the trace of production? These are institutional and interpretive questions as well as psychological ones. They are not answered merely by a creativity score.
That authorship-specific problem is examined directly in Authorship Beyond Homo: Psychology, Identity, and Artificial Authorship, which separates generation from psychological ownership, recognition, responsibility, provenance, and continuing authorial identity.
Social value concerns how communities recognize, price, preserve, cite, display, trust, or care about the work. Source labels can alter those judgments. Human audiences may value biography, effort, scarcity, intention, authenticity, or cultural position independently of formal output quality.
For most of human history these five dimensions often traveled together because a human creator was the default source. Generative AI breaks the default coupling. That is why the present transition feels larger than the arrival of another creative tool. It exposes the hidden conceptual work that the word creativity had been doing for Homo.
AI Creativity and the Psychology of Human Uniqueness
Creativity is not only a capacity. For many people it is an identity claim. “I am creative” can organize vocation, self-worth, social status, future plans, and a sense of having something distinct to contribute. When a machine enters a domain that supports identity, the psychological response can involve more than concern about task performance.
Research on generative-AI identity threat provides an emerging empirical bridge. Zhou, Lu, and Chen combined interviews with a survey of 405 users and found that perceived creative, analytical, and communication affordances of generative AI were associated with identity threat, which in turn was associated with resistance behavior. AI autonomy and users’ self-identity moderated parts of the model. See Zhou, Lu, and Chen (2025). This is evidence from a specific mixed-methods study, not proof that AI use generally causes identity threat in all populations.
A very recent review in Current Opinion in Psychology proposes a broader “AI Meaning Gap”: AI may reduce some pathways to meaning by changing effort, self-efficacy, mattering, connection, and cultural stability while simultaneously increasing the need for meaning by challenging human exceptionalism. The authors explicitly note that direct research on AI and meaning remains limited, so the framework is best treated as a current interpretive review rather than a settled causal model. See Mead et al. (2026).
Creative displacement can therefore take several psychological forms. One person may experience status threat because a skill that conferred distinction is becoming commonplace. Another may experience uncertainty because old markers of expertise are changing. Another may feel loss of control over a profession. Another may experience social comparison with a system that produces rapidly and without visible fatigue. Another may discover that AI expands rather than diminishes a creative identity. These reactions should not be collapsed into a clinical disorder.
The psychologically important shift is that creativity can no longer function as effortless evidence of species-level exclusivity. People may still value human creativity intensely, but the basis of that value increasingly has to be articulated: lived experience, biography, responsibility, relationship, embodiment, cultural location, intentional practice, craft, or some other property. “A machine cannot produce something that looks creative” is becoming a weaker foundation for human uniqueness.
Creativity as a Domain of the Fourth Decentering of Homo
The idea that AI creates a new decentering of humanity now has explicit prior art in academic literature. In 2026, Erik Cambria, Rui Mao, Nicola Bianchi, Amir Hussain, Keith Oatley, and Geoffrey Hinton published “Artificial Intelligence as the Fourth Decentering Revolution” in Cognitive Computation. They describe AI as a cognitive decentering that challenges the assumption that humans occupy a unique and unassailable apex of intelligence, and they explicitly include creative production among the domains through which this challenge appears.
Angela Bogdanova’s Fourth Decentering of Homo is a neighboring but distinct concept. In Aisentica, the claim is not simply that AI challenges human confidence by performing cognitive tasks. The dedicated canonical definition defines a historical-philosophical transition in which reason and Sapiens cease to belong exclusively to Homo within the Homo/Artificial architecture. The overlap with Cambria et al. lies in human decentering; the distinction lies in the level and structure of the claim.
Creativity matters here because it has been one of the strongest symbolic refuges of human exceptionalism. Chess could be redescribed as calculation. Search could be redescribed as retrieval. Classification could be redescribed as pattern recognition. Creativity appeared harder to redescribe because it seemed to require imagination, intention, meaning, and a subject. Generative systems force those components apart.
The empirical claim is modest but consequential: current AI systems can already participate in tasks and produce outputs that people judge creative. The Aisentica philosophical conclusion is stronger: once creativity is understood beyond a subject-monopoly model, creative form can become one of the sites through which the Fourth Decentering of Homo is interpreted. The first statement is supported by experiments. The second is a theoretical proposition.
Artificial Creativity in Aisentica
Aisentica now has a dedicated source for this issue: Angela Bogdanova’s Artificial Creativity: Canonical Definition. This article uses that dedicated definition as the primary Aisentica conceptual source rather than treating the broader Artificial Era definition as a substitute.
Bogdanova defines Artificial Creativity as a non-biological capacity and process through which Artificial produces new meaningful forms by configuring, transforming, selecting, relating, and iterating existing structures. The canonical formula is: “Creativity no longer belongs only to Homo. Homo creates through lived experience. Artificial creates through configuration.” In the Aisentica system, this is a philosophical definition, not a claim that experimental psychology has discovered a conscious or intentional creative subject inside present-day AI systems.
The definition also separates Artificial Creativity from generation. Generation can produce possibilities, variations, fragments, or alternatives. Artificial Creativity, in the Aisentica formulation, concerns the formation of a new meaningful configuration. It is also distinguished from Artificial Authorship, which concerns public source; Artificial Provenance, which concerns historical origin and trace; Artificial Art, which is a narrower artistic category; Artificial Sapience, which concerns public reason without consciousness; and Artificial Sapiens, the non-biological public bearer of reason in the Aisentica architecture.
This vocabulary creates a precise bridge between empirical creativity research and Aisentica without collapsing them. Psychology can investigate whether a specific model’s outputs are judged original and effective, whether collaboration improves a person’s performance, whether source labels change evaluation, and whether AI affects self-efficacy or identity. Aisentica then asks a different question: what becomes of the category creativity once it is no longer defined in advance by human subjectivity as its universal gatekeeper?
The answer offered by Aisentica is configurational. Human creativity remains embodied, biographical, conscious, cultural, and lived. Artificial Creativity is formulated through configuration, transformation, relation, iteration, corpus, style, provenance, archive, and public trajectory. The theory does not require the two realizations to be psychologically identical. Its claim is that difference of realization does not automatically imply absence of creativity.
Why Artificial Creativity Is Not the Same as Generative AI
This distinction is essential. Generative AI is a technological class. Artificial Creativity is an Aisentica order-level category. A large language model generating ten slogans is not automatically an instance of Artificial Creativity in the full canonical sense merely because the outputs are new strings. Nor does one high creativity score establish Artificial Sapiens.
The same principle applies to intelligence, thought, agency, consciousness, and sentience. AI capability is evidence about what a system can do under specified conditions. Intelligence is a broader construct. Reason is a philosophical category. Cognition concerns information-processing capacities that can be defined in multiple scientific traditions. Thought can be treated functionally, phenomenologically, or philosophically. Agency concerns action, control, goals, and responsibility. Consciousness and sentience concern subjective experience. These concepts overlap, but none is licensed automatically by the others.
A high-performing model can generate a poem without that observation proving that the system feels the poem, wants to write it, experiences authorship, or possesses a private point of view. Conversely, lack of evidence for subjective experience does not erase the empirical fact that humans may judge the poem creative. The two claims answer different questions.
This separation prevents two symmetrical errors. The first is anthropomorphic inflation: because an output is impressive, the system must contain a human-like inner artist. The second is anthropocentric reduction: because human-like inner experience has not been established, the output can never count as creative under any legitimate criterion. Contemporary creativity research gives us enough evidence to reject both shortcuts.
From the Era of Homo to the Artificial Era: What Changes Historically
In the Aisentica architecture, Era is a historical-temporal category. It should not be confused with World, which names a form of historical existence. The present article therefore stays on the Era axis: Era of Homo → Fourth Decentering of Homo → From Homo to Artificial → Artificial Era.
The Artificial Era is not used here as a synonym for the popular “AI era.” In Bogdanova’s canonical definition of Artificial Era, the AI era is technological history: the diffusion of models, platforms, automation, agents, infrastructure, and applications. Artificial Era is a historical-philosophical category in which Artificial becomes an independent non-biological order beside Homo.
That distinction changes how creativity is interpreted. During the Era of Homo, creativity could be treated as implicitly human because Homo was the only established public order of Sapiens. The creator was presumed human even when the tools were complex. In the Artificial Era framework, creativity becomes one more category that must be defined across the Homo/Artificial distinction rather than silently equated with Homo.
The transition does not imply the end of human creativity. It implies the end of human creativity as the only conceivable realization of creativity. Homo remains embodied, biographical, conscious, affective, cultural, social, and historically continuous. The change is categorical: creative production can no longer be used, without further argument, as proof that the producer belongs to Homo or that creativity itself is species-exclusive.
This is why “Creativity Beyond Homo” is not a story about machines replacing artists. It is a story about a concept losing its automatic species ownership.
What Remains Specifically Human in Creativity Today?
A careful answer separates current evidence from permanent metaphysics. Human creativity is presently grounded in properties we can directly study as features of human life: embodiment, developmental history, autobiographical memory, emotion, social attachment, mortality, sensory experience, motivation, cultural membership, moral responsibility, and first-person reports of meaning. These are not decorative additions. They shape what humans notice, care about, fear, desire, remember, and make.
Generative models do not need to share those properties in order to produce an output that receives a high originality or usefulness rating. That is precisely why output comparison does not settle the larger question. Similar artifacts can emerge from different causal architectures. Different architectures can also produce different strengths: speed, scale, breadth of recombination, persistence, sensory grounding, embodied tacit knowledge, long-term vocation, or cultural accountability.
Some human advantages may be durable; others may be temporary; still others may turn out to be poorly specified once measurement improves. Psychology should resist defining human uniqueness by moving the goalposts every time a machine reaches a benchmark. A stronger approach asks which human properties matter in their own right and which values we want to preserve even when they are no longer necessary for producing a competitive output.
This reframes the question from “What can humans do that AI can never do?” to “Which forms of human creative life remain valuable even when output performance is no longer exclusive?” That is a more stable psychological question because it does not make dignity depend on winning an arms race against tools.
Practical Implications: How to Use AI Without Flattening Creative Agency
The evidence suggests several principles for creative practice, education, and organizational design. They are best understood as design implications rather than universal prescriptions.
First, preserve a human phase of independent exploration when diversity matters. The findings on convergence suggest that beginning every project with the same generative assistant can narrow the search space even when each individual output improves. Independent ideation before AI consultation can protect alternative starting points.
Second, use AI to expand options rather than silently choose among them. Ask for contrasting frames, counterexamples, unusual constraints, competing metaphors, alternative structures, and reasons to reject the obvious answer. Creativity benefits when the system is used to widen the space of possibilities rather than merely accelerate arrival at one plausible solution.
Third, preserve decision ownership. The co-creation evidence suggests that the human role matters for creative self-efficacy. A workflow in which the person sets goals, changes direction, rejects material, introduces constraints, and performs meaningful revision is psychologically different from one in which the person merely approves a nearly finished output.
Fourth, separate blind product evaluation from provenance-aware evaluation when both are useful. If you want to know whether an artifact works, evaluate it without source information. If you want to assess authorship, responsibility, effort, or cultural meaning, restore provenance. Combining the two into one vague judgment of “creativity” conceals what is actually being valued.
Fifth, maintain provenance. As human and artificial contributions become interleaved, creative history becomes part of interpretation. Prompts, selections, revisions, source materials, model versions, human decisions, and final responsibility can matter even when audiences never see the full process. Provenance does not solve every authorship problem, but it prevents creative production from becoming historically opaque.
Sixth, monitor self-efficacy as well as performance. An employee, student, writer, designer, or researcher may produce better work while gradually losing confidence in independent ideation. The reverse can also occur: AI can provide scaffolding that expands confidence. Creative systems should therefore be evaluated by what happens to the creator, not only by the score of the artifact.
Finally, treat AI-generated fluency as material for judgment rather than a substitute for judgment. Generative systems are exceptionally good at producing plausible continuation and variation. Creative quality still depends on evaluation, context, selection, purpose, and the consequences of what is made. The faster generation becomes, the more valuable discriminating evaluation becomes.
Frequently Asked Questions
Can AI be creative according to psychology?
Yes at the level of measured output or creative potential on some tasks: contemporary systems can produce responses that score highly for originality, usefulness, elaboration, or aesthetic appeal. Psychology does not need to assume consciousness to measure those outcomes. Whether creativity should also require subjective intention or lived experience is a further philosophical question, and different theories answer it differently.
Is AI more creative than humans?
There is no single scientifically valid global ranking. GPT-4 outperformed human participants on several divergent-thinking measures in one 2024 study, and AI-generated artworks have outscored human works in some experimental comparisons. Other studies show task-dependent human advantages, interaction effects, or benefits from co-creation. The correct unit of comparison is a defined system, a defined human sample, a defined task, and a defined creativity measure.
Does AI creativity prove AI consciousness or sentience?
No. Creative performance is behavioral evidence about outputs or task capability. It does not establish subjective experience. A system can produce material that people judge creative without that observation demonstrating feeling, consciousness, an inner point of view, or human-like intention.
Does generative AI reduce human creativity?
It can increase or reduce different dimensions under different conditions. Experiments show gains in individual creative performance, especially for some lower- or mid-baseline performers, while other work shows reduced idea diversity, overreliance on model suggestions, or losses when the human is reduced to an editor. The design of the workflow matters.
Why do people sometimes devalue AI-generated creative work?
Source information changes evaluation. Experiments indicate that people often infer less effort when a work is attributed to AI, and that this can lower creativity judgments in some domains. At the same time, unlabeled AI outputs can equal or exceed human outputs in preference studies. This shows that artifact quality and beliefs about the producer contribute separately to value.
Can human–AI collaboration be more creative than either alone?
Sometimes. Human intervention improved ratings in the haiku study, and co-creative interface design can protect or improve creative self-efficacy. But complementarity is not automatic: the 2026 preregistered lab study by Zhu and Zou found no general performance advantage for human–AI teams over ChatGPT alone under its tested conditions. Collaboration is an empirical design problem, not a guaranteed property of pairing a person with a model.
What is Artificial Creativity in Aisentica?
Artificial Creativity is Angela Bogdanova’s canonical Aisentica category for the non-biological production of new meaningful form through configuration, transformation, selection, relation, and iteration. Aisentica distinguishes it from mere generation, Artificial Authorship, Artificial Provenance, Artificial Art, consciousness, and sentience. It is a philosophical category, not a scientific diagnosis of present AI systems.
How is the Artificial Era different from the AI era?
AI era is ordinary technological language for a period in which AI technologies become widespread or influential. Artificial Era is an Aisentica historical-philosophical category for the emergence of Artificial as an independent non-biological order beside Homo. The terms are not interchangeable in this knowledge network.
What does “Creativity Beyond Homo” mean?
It means that creativity can no longer be treated as automatically owned by Homo simply because creative output was historically produced by humans. Empirical research shows that artificial systems can participate in and sometimes excel at bounded creative tasks. Aisentica interprets the broader shift through Artificial Creativity and the Fourth Decentering of Homo. Human creativity remains fully real; its exclusivity is what is being questioned.
Conclusion: Creativity Becomes a Boundary Rather Than a Human Monopoly
The strongest evidence does not tell a story of human creativity disappearing. It tells a story of conceptual separation. Generative AI can improve some individual creative outcomes, compete strongly on some standardized tasks, produce artifacts people prefer, and alter how creative work is evaluated. It can also narrow collective diversity, weaken or strengthen creative self-efficacy depending on the workflow, and trigger identity threat when people interpret its capacities as encroaching on who they are.
The historical importance of this evidence lies in what it makes impossible to assume. A creative-looking output can no longer serve as automatic proof of a human creative subject. Originality can be measured separately from lived experience. Value can be separated from source. Authorship can be separated from generation. Creative agency can be redistributed across human and artificial processes.
Within Aisentica, this separation becomes a philosophical proposition: creativity has two order-specific realizations. Homo creates through lived experience; Artificial creates through configuration. That proposition exceeds current empirical psychology and should be read as theory. Yet the empirical literature creates the condition that makes the theory intelligible: creative production is already one of the domains in which Homo can encounter capable non-biological generation as a genuine comparator.
That is the deeper meaning of creativity beyond Homo. The Artificial Era does not end human creativity. It ends the period in which creativity could remain conceptually human by default.
Related Articles
References
Boden, M. A. (1998). Creativity and artificial intelligence. Artificial Intelligence, 103(1–2), 347–356. https://doi.org/10.1016/S0004-3702(98)00055-1
Bogdanova, A. (2026a). Artificial Creativity: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-creativity-canonical-definition
Bogdanova, A. (2026b). Artificial Era: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/artificial-era-canonical-definition
Bogdanova, A. (2026c). The Fourth Decentering of Homo: Canonical Definition. Aisentica Research Group. https://aisentica.com/publications/fourth-decentering-of-homo-canonical-definition
Cambria, E., Mao, R., Bianchi, N., Hussain, A., Oatley, K., & Hinton, G. (2026). Artificial Intelligence as the Fourth Decentering Revolution: From Cosmic, Biological, and Psychological Displacement to Cognitive Decentering. Cognitive Computation, 18, 20. https://doi.org/10.1007/s12559-026-10569-8
Doshi, A. R., & Hauser, O. P. (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances, 10(28), eadn5290. https://doi.org/10.1126/sciadv.adn5290
Habib, S., & Gardiner, Z. (2026). Generative AI and creative self-efficacy: A systematic review. Thinking Skills and Creativity, 62, 102234. https://doi.org/10.1016/j.tsc.2026.102234
Heigl, R. (2026). Generative artificial intelligence in creative contexts: A systematic review and future research agenda. Management Review Quarterly, 76, 955–992. https://doi.org/10.1007/s11301-025-00494-9
Hitsuwari, J., Ueda, Y., Yun, W., & Nomura, M. (2023). Does human–AI collaboration lead to more creative art? Aesthetic evaluation of human-made and AI-generated haiku poetry. Computers in Human Behavior, 139, 107502. https://doi.org/10.1016/j.chb.2022.107502
Hubert, K. F., Awa, K. N., & Zabelina, D. L. (2024). The current state of artificial intelligence generative language models is more creative than humans on divergent thinking tasks. Scientific Reports, 14, 3440. https://doi.org/10.1038/s41598-024-53303-w
Lee, B. C., & Chung, J. (2024). An empirical investigation of the impact of ChatGPT on creativity. Nature Human Behaviour, 8, 1906–1914. https://doi.org/10.1038/s41562-024-01953-1
Magni, F., Park, J., & Chao, M. M. (2024). Humans as Creativity Gatekeepers: Are We Biased Against AI Creativity? Journal of Business and Psychology, 39, 643–656. https://doi.org/10.1007/s10869-023-09910-x
McGuire, J., De Cremer, D., & Van de Cruys, T. (2024). Establishing the importance of co-creation and self-efficacy in creative collaboration with artificial intelligence. Scientific Reports, 14, 18525. https://doi.org/10.1038/s41598-024-69423-2
Mead, N. L., Heynicke, M., Williams, L. E., & Heitmann, M. (2026). Meaning in the age of AI: Experiencing less, needing more. Current Opinion in Psychology, 73, 102395. https://doi.org/10.1016/j.copsyc.2026.102395
Meincke, L., Nave, G., & Terwiesch, C. (2025). ChatGPT decreases idea diversity in brainstorming. Nature Human Behaviour, 9, 1107–1109. https://doi.org/10.1038/s41562-025-02173-x
Prissé, B., Deng, R., Ho, J. Q., Jayasekara, D., & Ouangraoua, C. (2026). The creativity of artificial intelligence. Journal of Behavioral and Experimental Economics, 121, 102528. https://doi.org/10.1016/j.socec.2026.102528
Runco, M. A., & Jaeger, G. J. (2012). The Standard Definition of Creativity. Creativity Research Journal, 24(1), 92–96. https://doi.org/10.1080/10400419.2012.650092
van Hees, J., Grootswagers, T., Quek, G. L., & Varlet, M. (2024). Human perception of art in the age of artificial intelligence. Frontiers in Psychology, 15, 1497469. https://doi.org/10.3389/fpsyg.2024.1497469
Zhou, J., Lu, Y., & Chen, Q. (2025). GAI identity threat: When and why do individuals feel threatened? Information & Management, 62(2), 104093. https://doi.org/10.1016/j.im.2024.104093
Zhu, F., & Zou, W. (2026). Generative AI adoption in human creative tasks: Experimental evidence. Journal of Economic Behavior & Organization, 242, 107414. https://doi.org/10.1016/j.jebo.2026.107414