Artificial-intelligence products are beginning to share a recognizable set of interface conventions, according to an analysis by web designer and writer Jim Nielsen. His survey describes an emerging “AI aesthetic” built from visual signals and interaction patterns that may outlast the current generation of chat products.
Nielsen frames the trend as part of a familiar design cycle. New technical constraints produce new interface ideas; some disappear, while others become general software conventions. He points to the hamburger menu as an earlier example. It spread during the growth of mobile interfaces, where limited screen space made a compact menu indicator useful, and later appeared in many other contexts.
In current AI products, the sparkle symbol has become a common shorthand for machine-assisted features. Rainbow treatments also appear frequently. These choices do not explain what a feature does, but they help vendors label an action as AI-related before a user opens it. Whether that symbolism remains useful as AI becomes less distinct from ordinary software is unresolved.
Other conventions arise from the underlying behavior of generative systems. Streaming text shows output arriving piece by piece and was refined in chat interfaces. Nielsen suggests that this pattern may remain closely tied to conversational AI because it represents generation in progress. Shimmering text has broader potential: although it can signal that a model is “thinking,” designers are also using it for asynchronous work such as fetching or computing.
The analysis also identifies subtler visual similarities. Nielsen compares desktop AI applications including Claude, Codex and Cursor with native macOS applications and observes that the AI products tend to use smaller, thinner icons. He associates the category with cream or beige palettes, orange accents and serif typefaces. These are observations about a sample of products rather than a formal survey of the market.
Interaction behavior may be becoming part of the aesthetic too. Nielsen describes controls that move when an interface redraws after a click, forcing the user to reposition the pointer. In his view, that instability echoes the nondeterministic character of AI output, although it can also create a frustrating user experience.
The analysis does not argue that every recurring choice will become a durable standard. Its central question is which patterns solve problems beyond AI chat. Streaming output may remain specialized, while shimmer effects or compact AI markers could migrate into ordinary applications. The answer will depend on whether users interpret these features consistently and whether designers preserve them after the novelty of the current AI wave fades more broadly.
