"Can you feel the vibes?": An exploration of novice programmer engagement with vibe coding CCS: Applied computing Interactive learning environments
Kiev Gama, Filipe Calegario, Victoria Jackson, Alexander Nolte, Luiz Augusto Morais, Vinicius Garcia · 2025
Read "Can you feel the vibes?": An exploration of novice programmer engagement with vibe coding CCS: Applied computing Interactive learning environments by Kiev Gama, Filipe Calegario, Victoria Jackson, Alexander Nolte, Luiz Augusto Morais, Vinicius Garcia (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-02750.txt.
Introduction
Kiev Gama and five co-authors study how beginners engage with vibe coding, building software by prompting AI. Instead of writing code directly, the programmer describes intent in natural language and steers the result. The style promises to open software building to far more people.
Its effects on learning, though, remain largely unexplored. Novices may gain confidence quickly, or may skip the struggle that builds real skill. The team's exploration of engagement offers early evidence for teachers. Schools need such evidence to decide how AI coding should enter the classroom. Prompt craft is a new skill, and how schools teach it will shape who gets to build with code at all.
Catalog record
| CREATOR | Kiev Gama, Filipe Calegario, Victoria Jackson, Alexander Nolte, Luiz Augusto Morais, Vinicius Garcia |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-02750
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
"Can you feel the vibes?": An exploration of novice programmer engagement with vibe coding CCS: Applied computing Interactive learning environments Kiev Gama, Filipe Calegario, Victoria Jackson, Alexander Nolte, Luiz Augusto Morais, Vinicius Garcia Abstract . Emerging alongside generative AI and the broader trend of AI-assisted coding, the term “vibe coding” refers to creating software via natural language prompts rather than direct code authorship. This approach promises to democratize software development, but its educational implications remain underexplored. This paper reports on a one-day educational hackathon investigating how novice programmers and mixed-experience teams engage with vibe coding. We organized an inclusive event at a Brazilian public university with 31 undergraduate participants from computing and non-computing disciplines, divided into nine teams. Through observations, an exit survey, and semi-structured interviews, we examined creative processes, tool usage patterns, collaboration dynamics, and learning outcomes. Findings reveal that vibe coding enabled rapid prototyping and cross-disciplinary collaboration, with participants developing prompt engineering skills and delivering functional demonstrations within time constraints. However, we observed premature convergence in ideation, uneven code quality requiring rework, and limited engagement with core software engineering practices. Teams adopted sophisticated workflows combining multiple AI tools in pipeline configurations, with human judgment remaining essential for critical refinement. The short format (9 hours) proved effective for confidence-building among newcomers while accommodating participants with limited availability. We conclude that vibe coding hackathons can serve as valuable low-stakes learning environments when coupled with explicit scaffolds for divergent thinking, critical evaluation of AI outputs, and realistic expectations about production quality. 1. Introduction Software is a part of everyday life, from critical infrastructure and finance to health, mobility, and communication. Understanding how it functions and being able to contribute to its creation can thus be perceived as a form of civic participation ( 41 ) . Starting to learn about software is difficult, though, as it requires specialized knowledge that novices must acquire before being able to produce non-trivial artifacts ( 42 ) . This expertise, in turn, is typically obtained through sustained education over longer periods of time, e.g., in the form of university courses. Sustained education, however, is not always feasible, particularly for individuals with limited time due to other commitments such as work and caregiving ( 22 ) . The recent emergence of Generative AI (GenAI) might help lower this barrier. Vibe coding in particular appears to be promising ( 4 ) . It allows basically anyone with access to suitable tools to create polished proofs-of-concept prototypes by simply describing what they want to build using natural language and iteratively refining what the tool produces ( 43 ) . It is, however, unclear at this point if people actually learn about technology while vibe coding. What is clear, though, is that novices will unlikely benefit from trying vibe coding alone. Pedagogy has long established that learning in a social environment together with like minded individuals and suitable support and guidance greatly benefits learning ( 13 ) . Hackathons can be a suitable setting in this context. They are time-bounded events during which participants form teams to collaboratively develop an artifact, which often takes the form of a piece of software ( 11 ) . Moreover, during hackathons, participants typically receive support in the form of mentors ( 29 ) . Prior research has shown the suitability of hackathons for educational purposes both in formal ( 36 ) and informal learning ( 28 ) settings. Hackathons have, however, also been criticized for being scary for newcomers to attend because of the assumption that technical proficiency is a requirement for attendance ( 15 ; 54 ) . It thus appears that running a vibe coding oriented hackathon can be mutually beneficial. Hackathons can provide a social environment where participants can try out and learn about technology while receiving support when needed and vibe coding can lower the barrier for attendance because it can lower the perception of technical proficiency being a prerequisite for attendance. In this paper, we report on an experience-based study of a one-day educational hackathon oriented toward vibe coding. Our goal was to understand how novices and mixed-experience teams utilize GenAI to transition from ideas to functional prototypes, what they learn in the process, and which event design choices facilitate inclusive participation. 2. Background and Related Work In this section, we will introduce the concept of vibe coding before discussing how GenAI and hackathons have been utilized in the context of software engineering (SE) education. 2.1. Vibe Coding Ongoing advances in AI, coupled with the natural language interfaces of AI-powered tools, such as Claude ( 8 ) , Replit ( 40 ) , Cursor ( 10 ) , and others, have led to the emergence of a new programming paradigm, colloquially known as “vibe coding”. First coined by the computer scientist Andrej Karpathy in a Tweet in February 2025, “vibe coding” is to “fully give in to the vibes, embrace exponentials, and forget that the code even exists…” ( 4 ) . Popularly, vibe coding has been interpreted as developing software solely by prompting an AI with no technical knowledge or expertise required ( 31 ; 43 ; 25 ) . Vibe coding has garnered significant attention within (e.g., ( 45 ; 25 ) ) and outside (e.g. ( 43 ; 50 ; 49 ) ) the SE community, with some noting that it can democratize the development of new applications, as applications can be easily developed without requiring coding experience ( 43 ) . Indeed, some consider vibe coding as a new development paradigm where humans and GenAI co-create software by conversing in natural language rather than code ( 27 ) or an evolution of first-generation AI-assisted programming ( 44 ) , which started with the availability of tools such as Copilot. Benefits of orchestrating and prompting AI-based tools to create an application include a lower barrier to entry for application development ( 27 ) , reduced time to market ( 27 ) , productivity benefits ( 25 ) , and greater autonomy for engineers ( 25 ) . Yet, risks have also been identified, including the potential for developers’ technical skills to degrade ( 27 ) , reduced code quality leading to issues such as security flaws ( 55 ) , and ambiguous authorship and accountability for the code ( 27 ) . Despite the hype surrounding vibe coding, empirical research into this emerging phenomenon is scarce. One study ( 17 ) , with CS and SE students of differing coding experience, found that more inexperienced students interacted via prompts more than advanced students, who noticeably reviewed and amended code more frequently. Also, the prompting styles varied with inexperienced students authoring more vague prompts compared to the advanced students, who wrote more specific and code-focused prompts. A study ( 44 ) of vibe coding live-streams notes that professional developers use an ecosystem of AI tools, and the objectives and requirements for the application evolve through developers’ interactions with the AI tools. Challenges faced include difficulties in communicating abstract ideas and understanding the capabilities and limitations of the AI tools. The study finds that technical expertise is beneficial in several ways, including identifying problems with the code, keeping the solution aligned with goals, and selecting the