What is it about?
This paper shares eight initial lessons learned during the early adoption of GenAI tools (ChatGPT and GitHub Copilot) by developers within a software division of a large Brazilian firm. One lesson learned included the need to adopt additional developer tools alongside GenAI ones; for example, code quality tools needed to be adopted to measure the potential impact of GenAI written code on the quality of the code base. Other lessons included the fact that ChatGPT is a powerful search companion and that Copilot struggles with legacy codebases. We also discuss the adoption approach taken by the company.
Featured Image
Photo by Fotis Fotopoulos on Unsplash
Why is it important?
GenAI is taking the software profession by storm, with individuals, teams, and organizations rapidly adopting GenAI developer tools such as GitHub Copilot and more general-purpose GenAI tools such as ChatGPT for reasons such as anticipated productivity improvements. Yet, there is little advice on how individuals and teams should adopt GenAI and what benefits they may experience or challenges they face. By sharing the benefits and challenges experienced by the individuals and teams contained in this study, other teams will profit from what we have learned before adopting GenAI tools.
Perspectives
This paper is an early insights paper and the first of a series of papers following the adoption of GenAI tools by developers within this company. I find it interesting that despite the rapid adoption of GenAI tools by developers, there are few studies based on "in the trenches" experiences that can help other organizations in (a) adopting such tools and (b) provide a dose of realism in the benefits and challenges they may see.
Victoria Jackson
University of Southampton
Read the Original
This page is a summary of: Early Results from a Study of GenAI Adoption in a Large Brazilian Company: The Case of Globo, January 2024, Springer Science + Business Media,
DOI: 10.1007/978-3-031-55642-5_13.
You can read the full text:
Contributors
The following have contributed to this page
