cross-functional-brainstorming

Source Metadata

Original cangjie-skill frontmatter from the distillation run:

name: cross-functional-brainstorming
description: |
  当用户需要启动AI项目的创意发现阶段，组织跨职能团队寻找AI应用场景时使用。
  用户在问"我们该做什么AI项目"、"怎么找到AI落地场景"、"brainstorm AI ideas"时激活。
  不适用于: 组织完全没有AI专业知识（无内外部顾问可用）、问题已明确定义无需创意探索的场景。
source_book: 《给所有人的AI入门课》 吴恩达 (Andrew Ng)
source_chapter: p14, p24
tags: [ideation, cross-functional, brainstorming, project-selection]
related_skills:
  - slug: automate-task
    relation: composes-with
  - slug: triple-due-diligence
    relation: composes-with

Cross-functional brainstorming — the intersection of AI experts × domain experts

R — Original Text (Reading)

"I often assemble teams that include AI knowledge experts and business domain experts to discuss together, so that we can jointly identify projects at the intersection of the two sets."

— Andrew Ng, p14

I — Methodological Skeleton (Interpretation)

High-quality AI project ideas come from the intersection of two circles: The first circle is "what AI can do" — this is mastered by AI experts, who understand the capability boundaries of current technology and the types of feasible tasks. The second circle is "what creates business value" — this is mastered by business domain experts, who understand which processes are costly, which steps are bottlenecks, and which improvements can bring the greatest returns. Having AI experts brainstorm alone will produce projects that are technically feasible but useless for the business; having the business team brainstorm alone will produce ideas that are valuable but technically infeasible. Only when the two types of experts sit together and look for projects at the intersection of "feasible" and "valuable" can they find AI application scenarios truly worth investing in. The core insight of this methodology is: AI project failures are often not technical failures, but rather "solving the wrong problem."

A1 — Application in the Book (Past Application)

Case 1: Call Center

Problem: How to use AI to improve call center efficiency
Use of the methodology: AI experts know that "automatic email classification" is technically feasible + domain experts know that "email handling is the largest source of workload in the call center"
Conclusion: Intersection project = automatic email classification and routing
Result: Became a classic entry point for AI adoption in call centers

Case 2: Radiology

Problem: How to help radiologists improve efficiency
Use of the methodology: AI experts know that "X-ray interpretation" is technically feasible + domain experts know that "reading images is the most time-consuming task for radiologists"
Conclusion: Intersection project = AI-assisted diagnosis for X-rays
Result: Medical imaging AI became the most active field of AI medical applications

A2 — Trigger Scenarios (Future Trigger) ★

In what situations will users need this skill?

An enterprise has just established an AI team and needs to find its first batch of AI landing projects
An organization wants to launch AI transformation but does not know which business scenario to start with
The technical team and business team operate in silos and need a collaboration mechanism to align on AI direction
A consultant is helping a client plan an AI strategy and needs a structured creative discovery process

Language signals (activate when these appear in the user's words)

"What AI project should we do"
"How do we find AI landing scenarios"
"brainstorm AI ideas"
"AI application scenarios"
"Where can AI help us"

Distinction from adjacent skills

With automate-task Difference: An automated task framework is "how to break down a direction into specific tasks," while cross-functional brainstorming is "who participates in idea discovery"—the former focuses on HOW, the latter on WHO.
and ml-feasibility Difference: The two-factor feasibility assessment evaluates whether a specific project is feasible, while cross-functional brainstorming is in the idea generation stage before a project is identified.
and ab-mapping Difference: A→B mapping formalizes the problem, while cross-functional brainstorming is creative exploration before formalization.

E — Executable Steps (Execution)

After the skill is activated, the agent should execute the following steps:

Identify AI expert resources
- Ask the user: "Do you have AI technical experts internally? If not, would you consider hiring external consultants?"
- Completion criteria: The user can clearly identify available AI expert resources (internal team or external consultants)
- Stop condition: If the user has no AI expert resources at all and cannot obtain them, it is recommended to solve the talent problem first before starting the project
Identify business domain experts
- Ask the user: "Which business lines best understand the pain points and value? Who are the senior experts in those domains?"
- Completion criteria: The user can list 2-3 key business domains and related experts
Organize a structured brainstorming session
- Suggest the user proceed with the following process:
  - Round 1: AI experts share "what AI can currently do" (list feasible task types)
  - Round 2: Business experts share "which areas are most valuable/most painful"
  - Round 3: Both sides jointly find the intersection and list candidate projects
- Completion criteria: The user obtains a list of candidate projects, each annotated with the two dimensions of "AI feasibility" and "business value"

B — Boundary (Boundary) ★

Do not use this skill in the following situations

The organization has no AI expertise at all and cannot obtain external consultant support (in this case, the talent problem should be solved first)
The problem is already clearly defined, and what is needed is an execution plan rather than creative exploration
There are severe departmental barriers within the organization, making true cross-functional collaboration impossible
The decision-making level already has a clear direction, and brainstorming is merely a formality

Failure modes the author warns about in the book

AI experts and business experts use different "languages," leading to low communication efficiency
Brainstorming generates a large number of ideas but has no subsequent evaluation mechanism, ultimately making implementation impossible
Domain experts have overly high or overly low expectations of AI capabilities, leading to bias in project selection

The author's blind spots / limitations of the era

The methodology assumes you can form a cross-functional team—in reality, departmental barriers, knowledge gaps, and power structures may hinder true collaboration
The book does not provide specific brainstorming facilitation techniques—in practice, structured guidance methods are needed
The methodology assumes AI experts and business experts can converse as equals—in reality, AI experts may dominate the discussion and suppress business perspectives
Does not discuss collaboration methods for remote/distributed teams

Adjacent methodologies that are easily confused

Difference from "design thinking": Design thinking focuses more on user experience and empathy, while cross-functional brainstorming focuses on matching AI technology with business
Difference from "agile development": Agile is a project management method, while cross-functional brainstorming is a project selection method

Related skills

automate-task (composes-with): After brainstorming to find a direction, use the task breakdown framework to break it down into specific executable tasks
triple-due-diligence (composes-with): After brainstorming to produce candidate projects, use triple due diligence for in-depth evaluation

Audit information

Verification passed: V1 ✓ / V2 ✓ / V3 ✓
Test pass rate: pending testing
Distillation time: 2026-06-29