Google is asking designers to follow their gut and let AI do the drawing. With its latest update, the company is shifting the creative process away from grids and guidelines toward something more instinctive, where a description of a feeling can be enough to start building a full-fledged interface.
This addition came about through Stitch, an experimental Google Labs tool that is shaping a new concept called “vibe design.” Instead of starting with schematics or component libraries, the process starts with intent. Users describe a desired mood, goal, or outcome, and the system translates that input into fully realized interface screens.
The workspace itself has been reimagined to fit this approach. Rather than functioning like traditional design software, Stitch acts more like an open canvas where text, images, links, and even code snippets can be added. An AI agent processes this context and generates multiple design paths, allowing users to explore variations without losing previous iterations.
The tool can also link pages to interactive flows, suggest how the user journey could unfold, and respond to changes in real time. Designers can request changes to layouts, colors, or navigation and see updates instantly, turning iteration into continuous interaction.
Google describes it as a natural extension of AI-driven workflows, similar to how “vibe coding” reframed software development based on prompts rather than syntax. Here, the same principle is applied visually, shifting the designer’s role from manually creating each element to managing and refining the results generated by the system.
The introduction allows Stitch to compete more closely with established design platforms like Figma and Adobe, while also broadening the interface design audience. By lowering the technical barrier, the tool opens up prototyping opportunities for developers, founders, and others who may not have a formal design background but can clearly articulate what they want to create.
At the same time, this approach raises familiar questions about authorship and originality. While the system can speed up early-stage work, the responsibility for shaping the final product still lies with the user.
Credits:
All images courtesy of Google.
