VibeX 2026 - 1st International Workshop on Vibe Coding and Vibe Researching

Vibe coding has emerged as a fashionable practice since early 2025, introduced by Andrej Karpathy, a computer scientist at Tesla, who described it as a new style of coding by typing prompts. It shifts developers’ attention from pure code production toward comprehensive capabilities through collaboration with artificial intelligence (AI). This transition is best understood through the AI Coding Spectrum, which tracks the evolution from simple token-level completions to autonomous AI Agents capable of multi-file edits and independent task execution.

Alongside this trend, a related notion, vibe researching, has appeared in gray literature, addressed by OpenAI’s Chief Scientist Jakub Pachocki and Chief Research Officer Mark Chen. It refers to a co-research approach in which research is conducted collaboratively with AI. While traditional Deep Research focuses on intensive, human-managed AI assistance to ensure rigor, Vibe Researching represents a paradigm shift toward agent-led workflows. In this model, autonomous AI Agents handle the heavy lifting of execution—such as literature synthesis, data analysis, and drafting—allowing the human researcher to focus on high-level vision, framing, and direction.

While vibe as a term originates from practice, its diffusion into research activities is expected to trigger broader discussions, practices, and potential paradigm shifts. It reflects the increasing presence and involvement of Agentic AI in software products and scientific knowledge production, where AI acts as a collaborator with human beings. This shift generates excitement but also provokes intense debate in academia: some researchers embrace the efficiency and exploratory capabilities of autonomous agents, while others express concerns about the loss of methodological transparency and academic integrity.

Given that vibe culture emerged first in the software industry, it is natural that the software engineering field will be among the earliest to experience its influence in research. The 1st International Workshop on Vibe Coding and Vibe Researching (VibeX 2026) provides a platform for scholars and practitioners to collectively explore the emerging vibe phenomenon through empirical evidence, theoretical discussion, and future visions. The workshop aims to attract a mixed audience of junior and senior researchers and practitioners, offering a timely, multi-perspective venue for examining how AI collaboration is reshaping software development practices and software engineering research.

Accepted Papers

Call for Papers

VibeX 2026 encourages contributions covering any topic related to collaborating with AI in software development and software engineering research. The workshop invites high-quality research papers that explore vibe coding and vibe researching from both theoretical and empirical perspectives, with a particular focus on the role of autonomous AI Agents.

Topics of Interest

Topics of interest include, but are not limited to:

• State-of-the-art and definition of vibe coding and vibe researching.
• The AI Coding Spectrum: Evaluating the transition from suggestive AI to fully autonomous development agents.
• AI Agents in Research: Architectures and workflows for autonomous literature review, hypothesis generation, and experimental execution.
• Deep Research vs. Vibe Researching: Comparative studies on methodological rigor, transparency, and the shift from “execution-heavy” to “vision-led” research.
• Tools and processes supporting vibe coding and vibe researching (e.g., Agentic IDEs, Researcher Copilot).
• Lessons from the successes and failures of vibe coding and vibe researching practices.
• Empirical studies on AI-assisted research, deep research, and vibe researching in software engineering.
• AI collaboration in literature review studies.
• Tensions in vibe coding and vibe researching: Speed vs. Quality and Rigor vs. Vibe.
• Vibe researching for junior and senior researchers.
• Ethical issues in vibe coding and vibe researching.
• Visions and conceptual models for vibe researching as a scientific methodology.
• VibeX, new AI-collaboration patterns emerging beyond vibe coding and vibe researching.

Submission Guidelines

Papers must be written in English, contain original unpublished work, and follow the official ACM Primary Article Template:

• https://www.acm.org/publications/proceedings-template

Deviating from the ACM formatting instructions may lead to a desk rejection. LaTeX users should use the following options:

\documentclass[sigconf,review]{acmart}
\acmConference[EASE 2026]{The 30th International Conference on Evaluation and Assessment in Software Engineering}{9–12 June, 2026}{Glasgow, Scotland, United Kingdom}

Papers must be submitted in PDF format through EasyChair:

• https://easychair.org/conferences/?conf=vibex2026

VibeX2026 will employ a single-blind review process. Accepted papers will be published in the joint workshop proceedings in the ACM Digital Library.

Paper Types

Authors may submit their papers under one of the following categories:

• Full research papers (max 10 pages, including references)
  Describing original and completed research (i.e., quantitative, qualitative, and mixed-methods studies) on topics related to collaborating with AI in software development and software engineering research.

• Ongoing research papers, vision papers, and positioning papers (max 5 pages, including references)
  Presenting preliminary results, conceptual explorations, methodological reflections, or visionary ideas that stimulate discussion and future research directions.

• Experience reports (max 5 pages, including references)
  Describing hands-on experiences, lessons learned, and reflective insights from collaborating with AI for coding and research in industrial environments.

Keynote Speaker

Prof. Tommi Mikkonen
University of Jyväskylä, Finland

Talk Title

The End of Vibe Coding as We Know It: Towards Systematic Practices in GenAI Native Software Engineering

Abstract

Vibe coding has gained a lot of interest in the software engineering community. Indeed, the ability to turn natural language presentations into runnable applications opens totally new opportunities for software developers as well as tor the wider public. However, vibes often lead to satisfying the immediate, straightforward use cases, and corner cases might be left unconsidered. In this keynote, I will start with the promise of vibe coding, and then advance to tools and techniques that are needed to support the development of trustworthy, dependable software with methods commonly associated with vibe coding.

Biography

Tommi Mikkonen is a professor of software engineering at University of Jyväskylä, Finland. He has a mixed background consisting of positions at different universities and in corporate research and development, where he has been prototyping new ways to design, develop, and deploy software in various environments. Recently, he has been leading two research projects on GenAI native software development, where the goal is to combine the best parts of AI supported and classical software engineering. During his career, Mikkonen has authored over 500 scientific publications and supervised over 500 theses of different kinds.