A selection of the most important recent news, articles, and papers about Coding with and for AI and Quantum Computing.
General News, Articles, and Analyses
1. Software expertise powers up quantum computing – Physics World
https://physicsworld.com/a/software-expertise-powers-up-quantum-computing/
(Wednesday, August 21, 2024) “Combining research excellence with a direct connection to the National Quantum Computing Centre, the Quantum Software Lab is focused on delivering effective solutions to real-world problems”
2. Top Programming Languages 2024 – IEEE Spectrum
https://spectrum.ieee.org/top-programming-languages-2024
Author: Stephen Cass
(Thursday, August 22, 2024) “At the top, Python continues to cement its overall dominance, buoyed by things like popular libraries for hot fields such as AI as well as its pedagogical prominence. (For most students today, if they learn one programming language in school, it’s Python.)”
3. Survey reveals the usage and perception of AI in the overall workforce, beyond just the realm of software development – SD Times
Author: Jenna Barron
(Thursday, August 29, 2024) “While many surveys focused on software developers have claimed that a majority of them are using AI to generate code, the story is different in the workforce overall.”
4. GenAI coding startup Magic lands $320M funding from ex-Google CEO and others — TFN
Author: Abhinaya Prabhu
(Friday, August 30, 2024) “Magic, an AI startup that creates models to automate code generation and software development, has landed $320 million in funding.”
5. How software developers feel about AI reshaping their work—and what this tells the rest of us
https://techxplore.com/news/2024-08-software-ai-reshaping-rest.html
“The narrative around the infiltration of artificial intelligence (AI) into our professional lives is often tinged with fear. Can a computer do what I do faster and more cheaply? Will my skills become obsolete? What if a robot takes my job?”
Technical Papers, Articles, and Preprints
1. [2408.14007] Using Large Language Models to Document Code: A First Quantitative and Qualitative Assessment
https://arxiv.org/abs/2408.14007
Authors: Guelman, Ian; Leal, Arthur Gregório; Xavier, Laerte; and Valente, Marco Tulio
(Monday, August 26, 2024) “Code documentation is vital for software development, improving readability and comprehension. However, it’s often skipped due to its labor-intensive nature. AI Language Models present an opportunity to automate the generation of code documentation, easing the burden on developers. While recent studies have explored the use of such models for code documentation, most rely on quantitative metrics like BLEU to assess the quality of the generated comments. Yet, the applicability and accuracy of these metrics on this scenario remain uncertain. In this paper, we leveraged OpenAI GPT-3.5 to regenerate the Javadoc of 23,850 code snippets with methods and classes. We conducted both quantitative and qualitative assessments, employing BLEU alongside human evaluation, to assess the quality of the generated comments. Our key findings reveal that: (i) in our qualitative analyses, when the documents generated by GPT were compared with the original ones, 69.7% were considered equivalent (45.7%) or required minor changes to be equivalent (24.0%); (ii) indeed, 22.4% of the comments were rated as having superior quality than the original ones; (iii) the use of quantitative metrics is susceptible to inconsistencies, for example, comments perceived as having higher quality were unjustly penalized by the BLEU metric.”
2. [2408.16601] Examination of Code generated by Large Language Models
https://arxiv.org/abs/2408.16601
Authors: Beer, Robin; Feix, Alexander; Guttzeit, Tim; Muras, Tamara; Müller, Vincent; Rauscher, Maurice; Schäffler, Florian; and Löwe, Welf
(Thursday, August 29, 2024) “Large language models (LLMs), such as ChatGPT and Copilot, are transforming software development by automating code generation and, arguably, enable rapid prototyping, support education, and boost productivity. Therefore, correctness and quality of the generated code should be on par with manually written code. To assess the current state of LLMs in generating correct code of high quality, we conducted controlled experiments with ChatGPT and Copilot: we let the LLMs generate simple algorithms in Java and Python along with the corresponding unit tests and assessed the correctness and the quality (coverage) of the generated (test) codes. We observed significant differences between the LLMs, between the languages, between algorithm and test codes, and over time. The present paper reports these results together with the experimental methods allowing repeated and comparable assessments for more algorithms, languages, and LLMs over time.”
