Context

"Vibe Coding" — a term coined by Andrej Karpathy (OpenAI co-founder) and named Collins Dictionary's Word of the Year 2025 — describes a new paradigm in AI-assisted software development where users express high-level intent in natural language rather than writing code directly. This dissertation examines how the term diffused and transformed across Platform X, drawing on Actor-Network Theory (Latour) and Science and Technology Studies.

Research Questions

What are the prevailing topics surrounding "vibe coding," and what are the identity characteristics of the groups driving each discourse? Which rhetorical strategies — modalities, citation of data, appeal to authority — do different stakeholder groups (developers, researchers, journalists, AI companies) use to legitimise their position? How does the discourse evolve through networked interactions, and how is meaning rhetorically strengthened or weakened over time?

Methodology

Corpus constructionBuilding a dataset from posts under #vibecoding on Platform X from February 2025 to early 2026, capturing high-frequency words, hashtags, user identity, reposts, comments, and engagement metrics.

Distant reading & frequency analysisApplying methods from Underwood et al. (2018) to identify dominant topics and rhetorical patterns across the corpus at scale using Python.

Social network analysisAdapting Elson et al.'s (2010) framework to extract conversational networks, categorise actors into stakeholder groups, and map their interactions via reposts, comments, and mentions using Gephi.

Topic visualisationAdapting Wilkens (2013) to visualise collective discourse patterns across the identified stakeholder groups.

Status

Expected completion and submission: June 2026. Full case study to be updated upon completion.