Towards AI-Native Software Engineering (SE 3.0): A Vision and a Challenge Roadmap

arXiv:2410.06107 · cs.SE, cs.AI · Submitted 2024-10-08 · Read on arXiv

cs.SE, cs.AI

Submitted: 2024-10-08

Updated: 2026-01-09

Code: https://github.com/Aider-AI/aider

Project page: https://lovable.dev

Terminology

• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Foundation Models (FMs) — 1× in this paper · explained in 2 episodes

Sources

• How is ChatGPT's behavior changing over time?
• Compiler.next: A Search-Based Compiler to Power the AI-Native Future of Software Engineering
• Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution
• Towards Conversational Development Environments: Using Theory-of-Mind and Multi-Agent Architectures for Requirements Refinement
• Agentic Software Engineering: Foundational Pillars and a Research Roadmap
• Rethinking Software Engineering in the Foundation Model Era: A Curated Catalogue of Challenges in the Development of Trustworthy FMware
• Speed at the Cost of Quality: How Cursor AI Increases Short-Term Velocity and Long-Term Complexity in Open-Source Projects
• Measuring Massive Multitask Language Understanding
• RouterBench: A Benchmark for Multi-LLM Routing System
• Measuring the Influence of Incorrect Code on Test Generation
• DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines
• Design choices made by LLM-based test generators prevent them from finding bugs
• SPICE: An Automated SWE-Bench Labeling Pipeline for Issue Clarity, Test Coverage, and Effort Estimation
• RouteLLM: Learning to Route LLMs with Preference Data
• Towards More Trustworthy and Interpretable LLMs for Code through Syntax-Grounded Explanations
• The Impact of AI on Developer Productivity: Evidence from GitHub Copilot
• Watson: A Cognitive Observability Framework for the Reasoning of LLM-Powered Agents
• LAB: Large-Scale Alignment for ChatBots
• Trae Agent: An LLM-based Agent for Software Engineering with Test-time Scaling
• Real-time Adapting Routing (RAR): Improving Efficiency Through Continuous Learning in Software Powered by Layered Foundation Models

Related papers

• Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
• GitSkills: A Dataset of Agent Skills on GitHub
• SABER: Benchmarking Operational Safety of LLM Coding Agents in Stateful Project Workspaces
• PackMonitor: Enabling Zero Package Hallucinations Through Decoding-Time Monitoring
• IntentCoding: Amplifying User Intent in Code Generation
• Incentives and Outcomes in Bug Bounties