Preliminary Abstract

In software engineering, the concept of “rubber-ducking”, where people verbally explain their problems to an inanimate object in order to externalise them, is a well-established debugging practice. While this can be helpful, traditional rubber-ducking remains entirely passive: the artifact does not respond, adapt, or provide feedback. We envision an AI-powered physical artifact that enables users to receive feedback in different stages of the debugging process and at different levels of detail through multimodal outputs, ranging from simple attention signals to actual help with coding problems. In this thesis, you will investigate the effect of different levels of feedback developers need from a rubber duck to support reflection and successful coding without disrupting their problem-solving process.

Goal

In this thesis, you will build a physical artifact in the form of a duck that supports voice input and offers multimodal forms of feedback. The system leverages existing AI technologies for software creation and debugging to generate situational feedback. This interface will then be evaluated in a user study (with CS students or developers) to find out how AI can support the problem-solving process in software engineering without completely taking over the process, therefore maintaining desirable properties like developer satisfaction, learning, and code understanding.

What you will do

• Find existing literature on software development and debugging with AI, human factors for software developers, and Tangible User Interfaces for software creation
• Design and implement a physical artifact that supports developer interaction for rubber-ducking
• Design and conduct an evaluation of your artifact
• Write a thesis documenting your process and its findings

What we expect

• You have experience with hardware prototyping (incl. working with microcontrollers, 3D printing, etc.)
• You have experience with software development, particularly AI-assisted software creation
• Solid skills in English reading and writing
• You can work and solve problems independently
• You are creative

What you get

• Two committed supervisors, weekly meetings, and hands-on advice
• Being part of state-of-the-art research on AI-user-interaction
• A bachelors/masters thesis