Working Programmer

Research on "the working programmer" focuses on improving the programmer experience through AI-assisted tools and understanding how these tools impact productivity. Current efforts center on evaluating large language models (LLMs) for code generation and repair, assessing their effectiveness in real-world scenarios beyond simple benchmarks, and developing metrics that better align with human programmer needs and perceptions of value. This research is crucial for advancing software engineering practices, enhancing programmer training, and ultimately improving the efficiency and reliability of software development.