Envisioning the Next-Generation AI Coding Assistants: Insights & Proposals

TL;DR

This paper shares insights from developing AI coding assistants, emphasizing clear expectations, integration, extendable design, and responsible data collection, while proposing future research challenges.

Contribution

It offers practical guidelines and open questions for creating advanced, responsible AI coding assistants based on real-world development experience.

Findings

AI assistants should set clear usage expectations

Integration with IDE features enhances usability

Extendable backend designs support scalability

Abstract

As a research-product hybrid group in AI for Software Engineering (AI4SE), we present four key takeaways from our experience developing in-IDE AI coding assistants. AI coding assistants should set clear expectations for usage, integrate with advanced IDE capabilities and existing extensions, use extendable backend designs, and collect app data responsibly for downstream analyses. We propose open questions and challenges that academia and industry should address to realize the vision of next-generation AI coding assistants.