Kasra Ferdowsi | Searcharxiv

arXiv2024

Validating AI-Generated Code with Live Programming

AI-powered programming assistants are increasingly gaining popularity, with GitHub Copilot alone used by over a million developers worldwide. These tools are far from perfect, however, producing code suggestions that may be incorrect in subtle ways. As a result, developers face a new challenge: validating AI's suggestions. This paper explores whether Live Programming (LP), a continuous display of a program's runtime values, can help address this challenge. To answer this question, we built a Python editor that combines an AI-powered programming assistant with an existing LP environment. Using this environment in a between-subjects study (N=17), we found that by lowering the cost of validation by execution, LP can mitigate over- and under-reliance on AI-generated programs and reduce the cognitive load of validation for certain types of tasks.

cs.HC↗

arXiv2023

The Usability of Advanced Type Systems: Rust as a Case Study

Advanced type systems that enforce various correctness and safety guarantees--such as linear and ownership types--have a long history in the Programming Languages research community. Despite this history, a human-centered evaluation of these type systems and their usability was all but absent, with empirical evaluations limited to testing their expressiveness in programs written by experts, i.e. the creators of the type system. In the past few years, this has begun to change with the adoption of a version of affine types and ownership in the popular Rust programming language. With the increase in Rust's popularity, various studies have begun empirically evaluating the usability of Rust's Ownership and Lifetime rules, providing a breadth of qualitative and quantitative information on the usability of such type systems. They found that despite Rust's general success in achieving its promise of safety and performance, these rules come with a steep learning curve and have been repeatedly cited as a barrier to adopting Rust. In this report, I provide a brief history of linear types and region-based memory management, which directly inspired Rust's type system. I then introduce Rust's Ownership and Lifetime rules, and present the state-of-the-art in academic research into their usability. I discuss both theoretical arguments and empirical evidence for why these rules are difficult to learn and apply, and survey existing work on addressing some of these difficulties. I also draw from broader works in the HCI and CS Education communities to recommend future work in this area.

cs.PL↗