Table of Contents
Introduction xxv
How to Use This Book xxxi
Part 1—Framing the Problem 1
Chapter 1 Case Study: How to Completely F*ck Up Your AI Project 3
A Boiling Pot of Spaghetti 3
Fail #1: Try to Replace a Trained Expert with AI 4
Fail #2: Forget About Cost vs. Benefit 4
Fail #3: No ML Training Data? No Problem! 5
Fail #4: It Makes No Difference What Question Your AI Model Is Answering 6
Fail #5: Don’t Worry About User Research—You Have an SME! 6
Final Thoughts 8
Reference 9
Chapter 2 The Importance of Picking the Right Use Case 11
Presuming That AI Will Be Telling Experts How to Do Their Job Is a Red Flag 11
Ask a Better Question 12
Reference 14
Chapter 3 Storyboarding for AI Projects 21
Why Bother with a Storyboard? 22
How to Create a Storyboard 24
Establishing Shot 24
Things 25
People 26
Faces 26
Transitions 27
Storyboard Conclusion 31
Storyboarding for AI 34
Final Thoughts 36
Design Exercise: Create Your Own Storyboard 36
Storyboarding Exercise Example: Death Clock 37
References 38
Chapter 4 Digital Twin—Digital Representation of the Physical Components of Your System 39
Digital Twin of a Wind Turbine Motor 39
The Digital Twin Is an Essential Modeling Exercise for Designing AI-Driven Products 43
How to Build a Digital Twin: An Example 44
Wait, There’s More! 46
Design Exercise: Create Your Own Digital Twin 48
Reflect 49
Design Exercise Example: Life Clock Digital Twin 49
Chapter 5 Value Matrix—AI Accuracy Is Bullshit. Here’s What UX Must Do About It 51
The Big Secret 51
Confusion Matrix: How Can Accurate AI Be Wrong? 53
Value Matrix: The AI Tool for the Real World 55
Training AI on Real-Life Outcomes to “Think” Like a Human 56
One More Example 58
Final Thoughts: The Importance of Human Cost/Benefit 58
Design Exercise: Create Your Own Value Matrix 59
Design Exercise Example: Life Clock Value Matrix 60
References 61
Part 2—AI Design Patterns 67
Chapter 6 Case Study: What Made Sumo Copilot Successful? 69
Strong Use Case 69
Clear Vision 70
Dedicated Full-Screen UI 71
AI-Driven Autocomplete 71
Next-Steps Suggestions 73
Final Words 74
References 74
Chapter 7 UX Best Practices for SaaS Copilot Design 75
The More Important the Task, the More Real Estate Is Required 75
Side Panel 75
Large Overlay 76
Full Page 77
Contents xix
SaaS Copilot Is Stateful 78
Specialized Fine-Tuned ChatGPT Model 78
Plug-Ins: Integrated Continuous Learning About Your Specific System 79
The IA of the AI Is Straightforward, Focused on Chat 80
Promptbooks: No Need to Twist into Pretzels to Write Prompts 82
Final Thoughts 82
Design Exercise: Create Your Own Mobile Copilot 83
Design Exercise Example: Life Clock Copilot 84
References 87
Chapter 8 Reporting—One of the Most Important Copilot Use Cases 89
Zoom AI Companion 89
Meeting Summary 89
Answer Questions About the Meeting 89
Set It and Forget It 90
UI Modality Switch 90
Microsoft Security Copilot 91
Executive Summary: A General Report 93
Pinboard: A Specialized Report Focused Only on Selected Key Details 93
Info for Report: Ignore Automatically vs. Pick Manually? 94
Security and Privacy 97
Design Exercise: Create Your Own Copilot Report 97
Design Exercise Example: Life Clock Copilot Report 98
Daily Report 98
Weekly Report 99
Chapter 9 LLM Design Patterns 103
Restating 105
Auto-Complete 106
Talk-Back 108
Initial Suggestions 108
Next Steps 111
Regen Tweaks 113
Guardrails 113
Design Exercise: Try Out the LLM Patterns 116
Design Exercise Example: “Life Copilot Plus” 117
Chapter 10 Search UX Revolution: LLM AI in Search UIs 121
The Current State of Search 121
Google Search 121
Amazon Search 122
The “Mysteries That Are Not Scary” Problem 122
Enter LLMs 126
Design Exercise: Design Your Own LLM Search UI 130
Design Exercise Example: Life Copilot LLM Search 130
Chapter 11 AI-Search Part 2: “Eye Meat” and DOI Sort Algorithms 133
What Are Dynamic Dashboards? 133
Beware of Bias in AI Recommendations 137
DOI: Degree of Interest/Sort Algorithms 138
Design Exercise: Create Your Own Dynamic Dashboards and Sort UI 143
References 143
Chapter 12 Modern Information Architecture for AI-First Applications 145
Design Pattern du Jour: The Canvas 145
Is Information Architecture Dead? 146
Amazon.com: Conventional Approach 148
“AI-Minus”? Homepage 148
Conventional Search Results Page 149
AI-Plus Item Detail Page 150
Conventional Maintenance Pages 151
AI-First Amazon.com Redesign 151
AI-First Analysis Overview Page 151
AI-First Category Analysis Pages 152
AI-First LLM Search 155
AI-First Item Detail 156
AI-First Maintenance Pages 159
Long Live Information Architecture! 159
Chapter 13 Forecasting with Line Graphs 165
Linear Regression 166
R-Squared 167
R vs. R-Squared 169
Forecasting with AI 169
Nonlinear Regression 170
Seasonality 172
Contents xxi
Forecasting an Aggregate Variable 173
Final Words 175
Design Exercise: Design Your Own Forecasting UI 175
Design Exercise Example: Life Clock Forecasting 176
References 177
Chapter 14 Designing for Anomaly Detection 179
Why Is Detecting Anomalies Important? 179
Four Main Anomaly Types 180
Point Anomaly 181
Change Point Anomaly 185
Contextual Anomaly 188
Curve Shape Anomalies 192
Design Exercise Example: Life Clock Anomaly Detection UI 195
References 196
Getting Ready for AI-pocalypse: Shorthand UX Design Notation as AI Prompt 197
Shorthand UX Design Notation 197
Shorthand Notation as AI Prompt to Go Directly to Working Code 198
What Is Human Work? 202
Chapter 15 UX for Agentic AI 203
What Are AI Agents? 203
How Do AI Agents Work? 204
Use Case: CloudWatch Investigation with AI Agents 205
Final Thoughts 209
References 211
Part 3—Research for AI Projects 213
Chapter 16 Case Study: MUSE/Disciplined Brainstorming 215
Design Idea #1 215
Design Idea #2 217
Design Idea #3 217
Design Idea #4 218
Design Idea #5 218
But Wait, Did You Catch That? 219
Design Exercise: Create Your Novel Designs Using Bookending 221
Design Exercise Example: Novel Design Ideas for Life Clock 222
References 222
Chapter 17 The New Normal: AI-Inclusive User-Centered Design Process 223
In the Beginning … 223
The Monkey or the Pedestal? 225
A New Way of User-Centered Thinking 226
What the Heck Is a Spike? 226
What Is the Role of Data? 227
Where Is the Customer in All This? 227
Why Is This Change Necessary? 227
How Does This Affect the Role of UX? 228
Does This Mean I Have to Learn About AI So That I Can Ask My Data Science Teammates Good Questions? 229
Final Handoff to Dev 229
Many More Changes to Come 230
Reference 230
Chapter 18 AI and UX Research 235
UX Techniques That Will Likely See Full Automation 235
Routine Usability Studies 235
Routine NPS Studies and Surveys 236
Collecting and Organizing the Research Data 236
Triangulation of Quantitative and Qualitative Insights 236
UX Techniques That Will Be Radically Augmented 236
Competitive Analysis 237
Identification of Novel Use Cases 237
RITE Studies 237
UX Techniques That Will Become Increasingly Valuable 237
Core Skills 237
Workshop Facilitation 238
Formative Research, Field Studies, Ethnography, and Direct Observation 238
Vision Prototyping 238
Augmenting the Executive Strategy 239
AI Bullshit 239
AI Strategic Analysis Tools That Replace Humans in Coming Up with Novel Ideas and Business Use Cases 239
AI Heuristics Analysis Replacing User Research and Design 240
AI Acting as “Synthetic Users” for the Purposes of Usability Research 240
Build Your Persona Using AI 240
Final Words 241
References 242
Contents xxiii
Chapter 19 RITE, the Cornerstone of Your AI Research 247
RITE Study vs. Usability Test 247
#1: RITE Studies Form the Core of the Design Process. Usability Tests Are Often Treated as QA 248
#2: RITE Studies Demand the Simplest Appropriate Prototypes That Change Rapidly. Usability Tests Often Mean Fancy Rigid Prototypes 249
#3: RITE Studies Produce Solutions. Usability Tests Produce Reports 250
A Fringe Benefit of RITE Studies 251
How to Conduct a RITE Study 252
A Few More RITE Rounds 255
The RITE Design Evolution 257
Dear Future: AI-Assisted RITE Methodology 257
Design Exercise: Run Your Own RITE Study 259
References 259
Part 4—Bias and Ethics 263
Chapter 20 Case Study: Asking Tough Questions Through Vision Prototyping 265
References 272
Chapter 21 All AI Is Biased 275
What Do You Expect When You Ask for “Biologist”? 275
How About “Basketball Player”? 275
Third Time’s the Charm: “Depressed Person” 277
References 281
Chapter 22 AI Ethics 283
Chapter 23 UX Is Dead. Long Live UX for AI! 299
AI Is Happening for Us, Not to Us 299
Staying on the Rollercoaster Is Optional 299
“UX Elitism” Is Over 300
Designers Are “Ambassadors of Innovation” 300
Core Skills Are in Demand 301
Combine Low-Fi UX Tools and Sophisticated AI Models 301
AI Is a “Wicked Problem” 301
AI Is Just Too Important to Be Left to Data Scientists 302
The Best AI is Augmented Intelligence 302
References 302
Index 303
