The widespread use of generative Artificial Intelligence (AI) has vastly transformed the software engineering landscape. Generative AI can massively write, review, explain and evolve code faster than developers can keep up with. It becomes materially impossible for developers to read, review, and understand the vast amount of AI generated code. Developers, however, face many risks: they must maintain and ensure a high degree of control over code quality and security, adapt and evolve their software engineering practices at the individual and organizational level, and avoid losing their software engineering expertise by delegating too much to AI agents.
The SAILS team falls under the field of software engineering. We research how to help developers navigate and master the rapidly evolving software engineering landscape. Our goal is to enable developers to (1) build a sufficient understanding of software systems that incorporate AI-generated code to make informed design and implementation decisions; (2) maintain control over key software properties, such as code quality, reliability, and security; (3) preserve and continue developing their software engineering expertise while benefiting from the productivity gains offered by generative AI; and (4) build effective software engineering organizations that foster productive human–AI collaboration. To do so, the SAILS team will place the augmented developer at the core of its research, investigating how developers adopt and collaborate with AIs, while designing tools, methodologies and practices that support the development and maintenance of secure, sustainable software systems.