Search papers, labs, and topics across Lattice.
This paper surveys 860 Microsoft developers to identify desired AI systems beyond code completion, revealing a demand for tools addressing the broader software development lifecycle. The analysis identifies 22 specific AI system concepts across five task categories, highlighting the problems they solve, their implementation challenges, and developer-imposed constraints. The key finding is a "bounded delegation" pattern: developers want AI to handle surrounding tasks but not core coding, emphasizing the importance of AI knowing where to stop.
Developers want AI to handle the grunt work around coding, but hands off when it comes to the creative core – revealing that the true value of AI tooling may lie in knowing where *not* to help.
Developers spend roughly one-tenth of their workday writing code, yet most AI tooling targets that fraction. This paper asks what should be built for the rest. We surveyed 860 Microsoft developers to understand where they want AI support, and where they want it to stay out. Using a human-in-the-loop, multi-model council-based thematic analysis, we identify 22 AI systems that developers want built across five task categories. For each, we describe the problem it solves, what makes it hard to build, and the constraints developers place on its behavior. Our findings point to a growing right-shift burden in AI-assisted development: developers wanted systems that embed quality signals earlier in their workflow to keep pace with accelerating code generation, while enforcing explicit authority scoping, provenance, uncertainty signaling, and least-privilege access throughout. This tension reveals a pattern we call"bounded delegation": developers wanted AI to absorb the assembly work surrounding their craft, never the craft itself. That boundary tracks where they locate professional identity, suggesting that the value of AI tooling may lie as much in where and how precisely it stops as in what it does.
Microsoft Research