in-IDE HAX — Human-AI Experience in Integrated Development Environment — names the field that studies the evolving dynamics of Human-Computer Interaction in AI-assisted coding environments. The term marks a shift of focus rather than a new tool: where Human-Computer Interaction asks how users interact with software and tools, Human-AI Experience is proposed for the case where AI is integrated into development workflows and acts as an active collaborator rather than just a tool, changing how developers work and interact with their environment.
Usage
The term is used by Human-AI Experience in Integrated Development Environments: A Systematic Literature Review, which argues that research in the area remains fragmented and sets out to give it a unified overview. That review makes the scope explicit by defining what counts as an IDE for the purpose: any workspace that simultaneously offers code editing, immediate execution or preview, and contextual AI feedback in the same window. Traditional desktop IDEs and notebook-style environments both satisfy all three criteria, as do plug-ins that embed a full editor and run code inline; tools that push AI suggestions outside the primary editing surface — the review’s examples are web explainers and offline static analysers — are excluded. The boundary is what separates in-IDE HAX from the broader question of how AI helps in software engineering: the focus is on what happens in the very place where code is written, inspected or tested.
The same review organises the field’s findings into three dimensions, derived through a staged qualitative coding procedure and carried over from an earlier survey by the same group:
- Impact — the effects of AI assistance on developers, tasks and workflows.
- Design — how AI is integrated into the IDE and interacted with.
- Quality — properties of AI outputs such as correctness, security and readability.
The review reports that its expanded corpus confirmed the adequacy of this three-part framing and added finer subscopes within each dimension without introducing any new top-level dimension. It also notes the practical consequence of studying experience rather than capability: a subset of the literature conceptualises programming with AI as a form of pair programming, treating the system as a collaborative partner, which the review reads as a move from isolated completions toward co-creative workflows and as raising concrete needs for training in model interaction and oversight in professional teams, and for scaffolds that keep learners in control in educational settings.
Related Terms
- AI IDE — the class of environment in which this experience takes place
- AI-Assisted Programming — the broader activity the field studies the experience of
- Automation Bias — one of the risks the field’s Impact dimension documents