Performative UI

Performative UI is an experimental project hosted on GitHub Pages by developer Li Zhang, exploring a new paradigm for user interfaces where the UI itself performs tasks rather than simply presenting controls for users to operate. The project is published at vorpus.github.io/performativeUI and appears to be a conceptual or research-oriented exploration rather than a production-ready commercial tool.

What It Is

Performative UI investigates the idea that modern AI-capable interfaces can shift from passive display surfaces to active agents — interfaces that "perform" on behalf of the user rather than waiting for explicit input. This positions the project within the broader conversation about agentic UI design, where the boundary between user action and automated execution becomes fluid. The project is authored by Li Zhang, who maintains a personal site and log at zha.ng.

Research and Conceptual Framing

The project appears to be a design research artifact or proof-of-concept, exploring how AI capabilities can be embedded directly into UI components to enable autonomous or semi-autonomous behavior. This kind of work sits at the intersection of human-computer interaction research and applied AI, examining what interfaces look like when they can act, not just display.

Audience and Use Case

Given its GitHub Pages hosting and experimental framing, Performative UI is primarily aimed at developers, designers, and researchers interested in the future of AI-native interfaces. It is not positioned as an enterprise product or end-user application, but rather as a conceptual contribution to the discourse around agentic and autonomous UI design patterns.

Current Status

The project is hosted as a static GitHub Pages site under the personal account of Li Zhang (vorpus). Based on available sources, it is a personal or independent research project with no commercial offering, versioned releases, or documented changelog. Its status reflects an exploratory, open-ended research artifact rather than an actively maintained software product.