Beyond the Canvas: Defining the "Invisible Skeleton" of AI Search Results

In the era of AI‑driven creation and decision‑making, design is shifting from "producing solutions" to "defining rules."
Traditional design models have focused on "delivering interfaces and workflows." However, as AI's capabilities in reasoning, generation, and judgment continue to advance, the role of the designer is no longer merely to produce one‑off solutions, but to establish a set of experience rules for human‑AI collaboration that can be executed over the long term. In our practice with Baidu App's search results, we have gradually observed:

The "results" users see are essentially generative products constrained by underlying rules.

The core of experience stability lies not in the visual style, but in the rule layer.

This workshop will center on the core proposition that "design thinking is about defining sustainable experience rules." We will explore how designers should re‑allocate the division of labor between humans and AI in a human‑AI collaborative working model. The session will cover the following main topics:

1. Understanding the Impact of Underlying Technological Changes in Search Strategy on Design
1.1 How search technology transformations affect the presentation of search results
1.2 How to anticipate the potential impact of technological changes on design

2. Design Methods and Case Study of Generative Search Top‑1 Results
2.1 Introduction to design methods: helping designers determine which rules must be controlled by humans and which generation and reasoning tasks can be delegated to AI
2.2 Practical case study: iterative evolution of generative search top‑1 result styles — a breakdown of the design thinking process

3. How to Scale Design Rules Systematically Across Multiple Platforms and Scenarios
Adaptation across multi‑device and cross‑scenario contexts

1、Design experience sharing
2、Workshop exercise: assembling a "generative search top‑1 result"
3、Hands‑on interaction: group presentations of design outcomes
4、Q&A session

1、User experience designers
2、Search product managers
3、Practitioners interested in AI

1、Understand the underlying architecture of generative search
2、Master methods for decomposing human‑AI task division
3、Learn to define rules for generative results
4、Build a "rule‑based thinking" design framework