AI as Tool, Not Muse: Student Views on Creativity in Physics

Author: Denis Avetisyan

A new study explores how students engaging with computational physics perceive the role of generative AI in their learning and creative processes.

Research indicates students primarily see generative AI as supporting individual skill development (mini-c creativity) but question its capacity for truly novel contributions to physics (little-c creativity).

While artificial intelligence increasingly permeates higher education, understanding student perceptions of its impact on fundamental cognitive processes remains crucial. This study, ‘Investigating student perceptions of creativity and generative ai in computational physics’, explores how physics majors conceptualize creativity and perceive the role of generative AI in their learning. Findings from essay-based interviews reveal that students primarily define creativity through individual exploration and skill development-manifesting as ‘mini-c’ and ‘little-c’ creativity-while viewing generative AI as a useful tool tempered by concerns regarding originality and accuracy. How might these evolving perceptions shape pedagogical approaches to fostering genuine innovation in STEM fields?

The Emergence of Creative Understanding

As generative artificial intelligence tools become increasingly prevalent, a nuanced understanding of how students perceive creativity is paramount. The capacity to innovate, problem-solve, and express original thought is no longer solely a human domain, prompting a re-evaluation of what constitutes creative work. Investigating student perspectives reveals not just what they consider creative, but also how they position themselves – and their own abilities – in relation to these new technologies. This is vital because differing perceptions can influence how students engage with AI – whether they see it as a collaborative partner, a competitive threat, or simply a tool for completing assignments. Ultimately, fostering a robust and informed understanding of creativity is essential for preparing students to navigate a future where human and artificial intelligence increasingly intersect.

Contemporary educational systems, while prioritizing cognitive skills, frequently neglect the explicit cultivation of diverse creative capacities. Traditional curricula often emphasize convergent thinking – arriving at a single correct answer – rather than divergent thinking, which encourages exploration of multiple possibilities. This oversight isn’t necessarily a deliberate omission, but rather a consequence of focusing on easily measurable outcomes and standardized testing. As a result, students may not receive dedicated instruction in creative problem-solving, idea generation, or the iterative process of experimentation crucial for innovation. The consequence is a potential stifling of individual creative expression and a narrowed understanding of what constitutes creative work, potentially limiting students’ ability to adapt and thrive in increasingly complex and rapidly changing environments.

Early investigations into student understandings of creativity suggest a prevalent, yet potentially limiting, focus on generating new or unusual ideas. This tendency often overshadows the crucial role of personal meaning and subjective experience in the creative process. While novelty is frequently highlighted as a defining characteristic, the intrinsic value students place on connecting with, and transforming experiences through their work is often undervalued. This emphasis on external validation – the ‘newness’ of an idea – may inadvertently discourage exploration of deeply personal, yet less outwardly ‘innovative’, creative endeavors. The research indicates a need to broaden the conceptualization of creativity to encompass not just the generation of novel outputs, but also the significance and resonance those outputs hold for the individual creator.

Recent investigation into the perceptions of creativity among six upper-division computational physics students reveals a dominant conceptualization rooted in the frameworks of ‘mini-c’ and ‘little-c’ creativity. This suggests that students largely define creativity not as groundbreaking, paradigm-shifting innovation – ‘Big-C’ creativity exemplified by historical figures – but rather as personally meaningful skill development (‘little-c’) and the everyday problem-solving expertise cultivated within a specific domain (‘mini-c’). The study indicates that these students prioritize the mastery of techniques and the incremental refinement of existing knowledge within physics, viewing creative acts as demonstrations of competence and nuanced understanding rather than radical departures from established norms. This perspective highlights a potential disconnect between conventional academic expectations and broader definitions of creative expression, suggesting a need to explicitly acknowledge and nurture diverse forms of creative thinking in STEM education.

Mapping the Landscape of Creative Acts

The Four C Model of creativity differentiates four distinct, yet related, forms. Mini-c represents nascent creativity experienced as personally meaningful insights and discoveries, typically within an individual’s immediate experience. Little-c denotes everyday creativity evident in problem-solving and adapting to life’s challenges; it’s demonstrable but not necessarily novel within a broader context. Pro-c is domain-specific expertise and skill development within a particular field, representing a level of mastery and professional accomplishment. Finally, Big-C signifies groundbreaking achievement recognized as transformative within a given field, often resulting in widely accepted new paradigms or impactful works; this level is historically rare and publicly visible.

The Four C Model differentiates between Little-c and Pro-c creativity based on the scope and validation of creative expression. Little-c creativity encompasses everyday problem-solving and personally meaningful creative acts that do not necessarily require external evaluation or recognition. In contrast, Pro-c creativity signifies domain-specific expertise and demonstrable skill within a defined field, typically involving evaluation by peers or authorities within that domain. This level requires a high degree of technical proficiency and often results in contributions that are recognized as novel and valuable by established experts, but may not yet reach the level of historical significance defining Big-C creativity.

The Four C Model facilitates a systematic assessment of student creative thinking by providing discrete categories – Mini-c, Little-c, Pro-C, and Big-C – against which individual conceptualizations can be measured. This allows researchers to move beyond generalized notions of creativity and instead identify specifically which levels of creative endeavor students prioritize or demonstrate understanding of. Analysis involves coding student responses – typically from narratives or interviews – for explicit references to, or implicit indications of, engagement with each ‘C’ level. The framework enables quantifiable data regarding the distribution of student perceptions across these levels, revealing patterns in how creativity is understood and valued, and allowing for comparisons between different student groups or educational contexts.

Analysis of student narratives within this study utilized the Four C Model as a coding framework, categorizing expressed concepts of creativity. Results indicated a predominant focus on Mini-c and Little-c creativity; specifically, students frequently described personal meaning-making and everyday creative acts. This suggests participants primarily conceptualized creativity as individually relevant experiences and practical problem-solving rather than domain-transforming achievements or demonstrated expertise. Coding involved identifying narrative segments aligning with each ‘C’ – Mini-c, Little-c, Pro-C, and Big-C – and quantifying the frequency of each category to determine the dominant conceptualizations of creativity present in the student sample.

Uncovering Creative Narratives

The study employed a qualitative research methodology centered on the collection of narrative essays from students to investigate their perceptions of creativity in relation to artificial intelligence. This approach prioritized in-depth understanding of individual viewpoints rather than numerical quantification. Students were prompted to compose essays detailing their personal definitions of creativity, experiences with AI tools, and reflections on how AI might influence creative processes. The resulting textual data provided a foundation for identifying prevalent themes and nuanced perspectives regarding the intersection of human creativity and emerging technologies, allowing for exploration of complex ideas not easily captured through quantitative methods.

The use of narrative essays as a primary data collection method prioritized student agency and allowed for the capture of complex perspectives beyond the limitations of structured questionnaires. This approach yielded detailed, first-person accounts of student experiences and beliefs regarding creativity and artificial intelligence, providing a depth of information not typically obtained through quantitative methods. The open-ended format encouraged participants to articulate their thoughts in their own terms, resulting in rich, nuanced data that captured the subtleties and individual variations in their perceptions. Consequently, the resulting corpus of essays provided a robust foundation for identifying emergent themes and patterns related to student understanding of creativity.

Coding analysis of the student essays involved a multi-stage process to identify themes aligning with the Four C Model of Creativity – Curiosity, Creativity, Critical Thinking, and Communication. Initially, a research team independently reviewed a randomly selected subset of essays to establish a preliminary codebook based on indicators of each ‘C’. This codebook was then iteratively refined through team discussion and testing on additional essays to ensure inter-rater reliability. Subsequently, all essays were coded by trained researchers, with instances of each ‘C’ indicator being flagged and categorized. Quantitative data, including the frequency of each code, was compiled and analyzed to identify dominant themes and patterns. Furthermore, qualitative analysis of coded segments provided contextual understanding of how students demonstrated these creative capacities in relation to AI technologies.

The coding analysis of student narratives revealed specific patterns in how students define creativity when considering emerging technologies. Themes consistently linked to the Four C Model – Creativity as Cognitive, Combinatorial, Conceptual, and Experiential – were identified and quantified across the essay corpus. This allowed researchers to move beyond broad definitions of creativity and pinpoint the specific cognitive processes, combinational thinking, conceptual blending, and subjective experiences students associate with AI and other new technologies. The frequency and interrelation of these coded themes provided a detailed map of student conceptualizations, demonstrating how they perceive both the opportunities and challenges AI presents to creative endeavors and personal expression.

The Evolving Landscape of Creative Agency

Research indicates a pronounced tendency among students to anthropomorphize artificial intelligence, particularly when interacting with generative models like ChatGPT. This isn’t merely a matter of conversational ease; students frequently ascribe qualities of creative agency – intention, originality, and even emotional depth – to these algorithms. The observed phenomenon suggests a cognitive inclination to perceive intelligence and creativity as inherently linked to a conscious entity, leading individuals to project these attributes onto AI despite its fundamentally different operational basis. This predisposition towards humanization carries implications for understanding how students evaluate AI-generated content and potentially impacts their own creative processes, necessitating a focused effort on distinguishing between algorithmic output and genuine human expression.

The increasing sophistication of artificial intelligence demands a renewed emphasis on critical thinking, particularly concerning the discernment of authentic creative work from AI-generated outputs. While generative models can produce novel content, attributing genuine creativity requires evaluating the underlying process – intention, emotional resonance, and personal meaning-making – qualities currently exclusive to human cognition. Educational efforts must therefore move beyond simply identifying AI-created content and instead focus on developing students’ abilities to analyze, evaluate, and ultimately define creativity itself, fostering a deeper understanding of what constitutes original thought and artistic expression in an age where imitation and algorithmic generation are increasingly prevalent. This analytical skillset is not merely about detecting artificiality, but about strengthening the very foundations of human creative capacity.

Effective educational programs must move beyond simply assessing demonstrable innovation and instead cultivate creativity across its entire spectrum. Recognizing that creative expression isn’t solely about groundbreaking achievements, curricula should prioritize nurturing ‘Mini-c’ – the universally accessible capacity for personal meaning-making through everyday experiences. This foundation then allows for the development of ‘Pro-C’, or professional creativity demonstrated through expertise within a specific field, and ultimately ‘Big-C’ creativity, which manifests as truly novel and impactful contributions to culture or knowledge. By intentionally fostering these tiered levels, educators can empower all students to not only utilize emerging technologies like AI but also to confidently express their unique perspectives and contribute meaningfully to an increasingly complex world, ensuring creativity remains a uniquely human endeavor.

Recent investigations reveal a nuanced student perspective on generative AI, positioning it primarily as a facilitator for self-directed study rather than a source of original thought. While students readily acknowledge the potential of tools like ChatGPT to aid in research and information gathering, a consistent skepticism surrounds its capacity for genuine creativity and consistently reliable output. This suggests that effective educational integration shouldn’t prioritize simply using the technology, but cultivating a critical awareness of its strengths and limitations. A balanced approach is crucial, one that leverages AI’s supportive capabilities while simultaneously reinforcing the unique value of human ingenuity, critical analysis, and independent thought processes within the learning environment.

The study illuminates how students perceive generative AI not as a source of novel insight, but as a powerful extension of individual exploration-a means to enhance ‘mini-c’ creativity through personalized learning. This resonates with a sentiment echoed centuries ago by Isaac Newton: “If I have seen further it is by standing on the shoulders of giants.” Just as Newton acknowledged building upon prior knowledge, students see these tools as scaffolding for their own understanding, accelerating personal progress within established frameworks. The research suggests that while AI can augment existing skills, the leap to genuinely original thought – ‘little-c’ creativity – remains firmly within the realm of human endeavor, a point where accumulated knowledge meets inspired intuition.

What’s Next?

The observed student perspective-AI as a proficient assistant for personal exploration (mini-c) but a questionable originator of novel physics insights (little-c)-reveals a pragmatic boundary. It isn’t a dismissal of the technology, but a tacit assertion of what constitutes genuine creative work. The study highlights how easily tools are assimilated into existing cognitive frameworks; small decisions by many participants produce global effects. Attempts to control the definition of creativity, or to mandate originality, seem increasingly futile. The interesting question isn’t whether AI can be creative, but how its presence reshapes student understanding of creativity itself.

Future inquiry should move beyond simply assessing perceptions. Examining how students actually use generative AI in problem-solving-the specific prompts, the iterative refinement, the critical evaluation of outputs-will offer a richer dataset. Furthermore, longitudinal studies are needed to determine whether early skepticism gives way to a more nuanced understanding of AI’s potential, or solidifies into a persistent view of it as a sophisticated, but ultimately derivative, instrument.

The Four C model, while useful for categorization, may prove too rigid. Perhaps creativity isn’t a hierarchy, but a complex, interconnected web. Rather than seeking to ‘unlock’ little-c creativity through AI, it may be more productive to investigate how AI facilitates, or hinders, the development of mini-c and little-c capabilities-acknowledging that influence, not control, is the operative principle.

Original article: https://arxiv.org/pdf/2603.12154.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/

The Emergence of Creative Understanding

Mapping the Landscape of Creative Acts

Uncovering Creative Narratives

The Evolving Landscape of Creative Agency

What’s Next?

See also: