GitHub Copilot AI pair programmer: Asset or Liability?
By Arghavan Moradi Dakhel, Vahid Majdinasab, Amin Nikanjam, Foutse Khomh, Michel C. Desmarais, Zhen Ming (Jack) Jiang. 2022
509 citations recorded by the source corpus.
Automatic program synthesis is a long-lasting dream in software engineering. Recently, a promising Deep Learning (DL) based solution, called Copilot, has been proposed by OpenAI and Microsoft as an industrial product. Although some studies evaluate the correctness of Copilot solutions and report its issues, more empirical evaluations are necessary to understand how developers can benefit from it effectively. In this paper, we study the capabilities of Copilot in two different programming tasks: (i) generating (and reproducing) correct and efficient solutions for fundamental algorithmic problems, and (ii) comparing Copilot's proposed solutions with those of human programmers on a set of programming tasks. For the former, we assess the performance and functionality of Copilot in solving selected fundamental problems in computer science, like sorting and implementing data structures. In the latter, a dataset of programming problems with human-provided solutions is used. The results show that Copilot is capable of providing solutions for almost all fundamental algorithmic problems, however, some solutions are buggy and non-reproducible. Moreover, Copilot has some difficulties in combining multiple methods to generate a solution. Comparing Copilot to humans, our results show that the correct ratio of humans' solutions is greater than Copilot's suggestions, while the buggy solutions generated by Copilot require less effort to be repaired.
Subfields: Information Systems, Software, Code Generation & Program Synthesis
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Information Systems
- Paul Denny
- Juho Leinonen
- Arto Hellas
- Sami Sarsa
- Siddharth Garg
- Ramesh Karri
- Hammond A. Pearce
- Andrew Luxton-Reilly
In Software
- Paul Denny
- Juho Leinonen
- Arto Hellas
- Sami Sarsa
- Brendan Dolan-Gavitt
- Andrew Luxton-Reilly
- Siddharth Garg
- Zhi Jin
In Code Generation & Program Synthesis
- Paul Denny
- Juho Leinonen
- Sami Sarsa
- Arto Hellas
- Brendan Dolan-Gavitt
- Siddharth Garg
- Zhi Jin
- Ramesh Karri
Citations in the corpus
- Lost at C: A User Study on the Security Implications of Large Language Model Code Assistants (2022, background): “These LLMs can write functionally correct code, with studies showing capabilities in solving introductory programming tasks [17, 18] and algorithmic challenges [19]).”
- The Metacognitive Demands and Opportunities of Generative AI (2023, background): “During prompt formulation , the open-endedness of many current prompting interfaces requires users to have self-awareness of their specific task goals, and be able to decompose their tasks into smaller sub-tasks so as to verbalize these as effective prompts (Figure 2a).”
- Programming Is Hard - Or at Least It Used to Be: Educational Opportunities and Challenges of AI Code Generation (2022, background): “In this work we do not discuss wider societal (e.g., economic [8], political [17]) considerations presented by AI-generated code, nor those specific to advanced/professional programmers (e.g., [9, 31]).”
- Generative artificial intelligence: a systematic review and applications (2024, unknown): “In the field of software engineering, generative AI is introduced to help write better code and solve errors in the code, debugging and even writing the documentation of the work (e.g., Copilot) [92].”
- CodeAgent: Enhancing Code Generation with Tool-Integrated Agent Systems for Real-World Repo-level Coding Challenges (2024, methodology): “2 Background GPT-3 (Brown et al., 2020) models on code corpus to produce CodeX (Chen et al., 2021) and GitHub Copilot (Dakhel et al., 2023).”
- Computing Education in the Era of Generative AI (2023, background): “Other recent work has shown that these tools can reliably generate correct code for common algorithms such as insertion sort and tree traversal [9], and for problems that are typical in more advanced computing courses focusing on data structures and algorithms [16].”