’Read the World and Read Your Mind‘: Ripple as AI Native UI Interface
- Yihua Li
- Nov 23
- 5 min read

11/23/2025
Cory Yihua Li - Echuu Research
Live Demo: ripple-ui-beta.vercel.app
Imagine walking into a room. You see a lamp. You don't need to open a menu to find the “Turn On” command. You don't need to download a “Lamp Controller App.” You simply reach out, touch it, and your intent—to illuminate the room—is realized.
For 40 years, our digital lives have been trapped in a Tool-Centric paradigm. Whether in Photoshop or a CRM, the flow is rigid:
Learn the Menu → Find the Tool → Apply to Object.I built RippleUI to break this sequence.I am introducing an Object-Centric paradigm where the interface doesn’t exist until you touch the digital world.
Its mission is two fold: Read the World and Read Your Mind.
1. Read the World: From Pixels to Semantic Affordance
In traditional GUIs, a click is just a coordinate (x, y). The computer has no idea what you meant. RippleUI changes this with Semantic Raycasting.
Beyond the Bounding Box: Using advanced vision encoders like Gemini 2.0 Flash, we instantly ground your touch in reality. We don't just see a selection box; we see the story.
"User tapped a vintage vending machine on a neon-lit street."
Semantic Affordance: Even in 2D, the image is no longer flat pixels. It becomes a field of interactable meaning. The system recognizes the object's category, state, and environment, preparing it for intelligent manipulation.
1.读懂世界:从像素到语义可供性 在传统 GUI 中,点击只是屏幕上的一个坐标 (x, y),系统并不理解你碰到了什么。 RippleUI 通过 Semantic Raycasting (语义射线) 改变了这一点。
超越框选:利用 Gemini 2.0 Flash 强大的视觉能力,我们能瞬间“锚定”你的触碰。系统看到的不再是冷冰冰的像素,而是具体的场景:
“用户触碰了霓虹灯街道上的一台复古自动售货机。”
语义可供性 (Semantic Affordance):即使在 2D 屏幕上,图像也不再是死板的背景。每一个物体都被 AI 赋予了意义,处于一种“待命”状态,随时准备响应你的交互。
2. Read Your Mind: The Cognitive Bridge
Once RippleUI knows what you touched, it infers why.
No tool-hunting.No menu-digging.
RippleUI reverses the flow:The tool comes to you.
Probabilistic Intent:Based on object + context:
Snowy street + sneaker → “Winter-proof this”Living room + vintage radio → “Play music”
While the AI thinks, you see a ripple expanding from your fingerprint.This is a Cognitive Bridge: turning AI latency into a moment of anticipation and meaning.
2.读懂你:认知桥梁
知道你触碰了什么之后,更重要的是理解你为什么触碰。 RippleUI 在这里彻底反转了交互逻辑: 不再是你找工具,而是工具来找你。
基于概率的意图推断:我们根据“物体 + 上下文”来预判你的需求:
雪地 + 运动鞋 → “生成冬季适配版”
客厅 + 老式收音机 → “播放音乐”
认知桥梁 (The Cognitive Bridge):在 AI 思考的毫秒间,你会看到从指纹处荡漾开的水波纹。这不只是动效,它是一座桥梁:把等待 AI 推理的“延迟”转化为一种充满意义的“期待”,连接了你的直觉与 AI 的行动

Why a Fingerprint and Ripples?

The fingerprint marks your exact touch point.The ripple visualizes how RippleUI interprets your intent.
From the center outward, each ring represents a ranked hypothesis—from the most likely action to more exploratory options. As the rings expand, RippleUI turns those predictions into actionable buttons: edit, compare, transform, generate, inspect.
Instead of searching for tools, you touch the world and the interface unfolds—intent revealed from the inside out.
3.为什么是指纹 + 水波纹?
指纹代表你的触点。水波纹代表 RippleUI 的推理过程。
从中心向外,每一圈都是一层意图排序:中心是最可能动作,外圈是更探索性的选项。随着扩散,每一圈都会被转化成可点击的按钮:编辑、比较、生成、变形、分析。
你不需要找工具。你触碰世界,界面从你的指纹中心向外展开——意图从内向外被揭示。
在Echuu的设定和世界观里,水和天空的意象是核心,我希望通过这个设计体现用户与这个界面交互的作用力
4. The Interface Follows: Ephemeral Apps
If your intent is “compare cameras,” why should the AI return a paragraph?
Drawing on Google Research’s Generative UI, RippleUI believes the interface should be generated, not pre-installed.
Instant Engineering Teams:For each interaction, the AI acts as a mini product team (PM + Designer + Engineer) to build a tailored UI.
Dynamic Widgets:The ripple can evolve into:
A rotatable 3D diagram
A recommendation panel
A flight comparison module
The “app” becomes temporary.Interfaces emerge when needed, dissolve when done.
4. 界面随后生成:瞬时应用(Ephemeral Apps)
如果你的意图是“比较相机”,为什么界面要给你一段文字?
基于 Google 的 Generative UI 研究,RippleUI 相信界面应该是生成的,而不是预置的。 research.google
瞬时工程团队:AI 会像一个小型产品团队(PM + 设计 + 工程)一样,为你即时生成对应界面。
动态组件:水波纹可能变形为:
可旋转的 3D 图解
推荐模块
机票比较工具
App 不再是静态容器。界面只在需要时出现,用完即消失。
Why RippleUI Matters for Echuu
Echuu explores what AI-native presence looks like: characters, worlds, interfaces that behave like living environments.
RippleUI is the interaction layer of that vision:
The fingerprint = personal identity
The ripples = shifting intention
The world = responsive, alive
Touch → Interpretation → Response.This is the logic that underpins Echuu’s entire system.
I want to explore the possiblity to use body as interface to unlock the possibility of Vtuber world , ripple UI is a start point and will expand this concept in future streaming context.
RippleUI 为什么属于 Echuu 的一部分?
Echuu 探索的是 AI 原生存在:角色、世界、界面,都像活的生态一样回应你的触碰。
RippleUI 是这个愿景的交互层:
指纹 = 用户身份
水波纹 = 意图扩散
环境 = 会回应的世界
触碰 → 系统理解 → 系统回应这就是 Echuu 世界的底层逻辑。
Join the Experiment
RippleUI v0.0.1 is powered by Gemini 2.0 Flash and React 19.We are refining the Cognitive Bridge and intent engine before rolling out full 3D and Generative UI features.
We invite developers, designers, and visionaries to touch the future of interfaces.
Live Demo: ripple-ui-beta.vercel.app
Stop typing prompts.Start touching the world.
加入实验
RippleUI v0.0.1 使用 Gemini 2.0 Flash 和 React 19 构建。我们正在完善认知桥与意图引擎,之后将开放 3D 与生成式界面能力。
欢迎开发者、设计师、研究者加入未来界面的尝试。
Live Demo: ripple-ui-beta.vercel.app
停止输入 Prompt。开始触碰世界。
References / 参考资料
Yaniv Leviathan, Dani Valevski & Yossi Matias. Generative UI: A rich, custom, visual interactive user experience for any prompt. Google Research Blog, November 18 2025. research.google
Ryo Lu. AI Interaction Logic Shift – Talk. YouTube, video ID: PQhcHrCyU8M. (探索AI 原生 UI/UX 形态)


Comments