Design and AI innovation
Summary
What we design is changing; therefore, how we design is also changing. Design innovation is being affected by emerging AI technologies. In this talk, I will set the context for the role of design in creating purposeful and pragmatic technology, both historically and today. I will talk about some of the problems with AI innovation and I will show some examples from our research showing the impact of design in creating, developing, and deploying AI systems, with the goal of creating better social systems, better economic relations, and a better world in which to live.
Key Insights
-
•
AI is transforming virtually every aspect of design, requiring new methods, processes, and roles for designers.
-
•
There is a significant innovation gap in AI products: most fail due to poor data, no customer value, lack of desirability, or ethical risks.
-
•
Designers often join AI product development too late, after data and models are chosen, limiting innovation and ethical integration.
-
•
The economy is shifting from manufacturing products to delivering integrated product-service ecologies, exemplified by companies like Starbucks and Walmart.
-
•
Service design requires a multi-stakeholder, holistic view that accounts for all affected parties, beyond obvious users.
-
•
AI should be treated as a design material, with specific capabilities and features serving as design patterns for innovation.
-
•
Digital algorithmic management in hospitality often strips workers of autonomy and requires new designs that restore agency and transparency.
-
•
Moderate AI performance, not perfect AI, often provides the best balance of value and usability in AI products, e.g., voicemail transcription.
-
•
Ethical concerns like privacy, bias, and surveillance are often unmanaged in AI product teams, highlighting a critical need for explicit ownership.
-
•
Collaborations with unions and workers can lead to meaningful changes in AI deployment, improving worker wellbeing and influencing policy.
Notable Quotes
"I call myself an accidental vagrant because I come from design but work in computer science—designers now live in contexts they haven't traditionally been."
"AI is poised to have a similar impact as electricity, transforming almost everything in the coming years."
"If we deploy AI too quickly without foresight, AI will burn in ways we cannot control."
"You would never implement an algorithm on a doctor without speaking to them, but in hospitality, it’s done to workers all the time."
"Most designers join AI projects after the data has been discovered and the model built—that’s far too late for meaningful innovation."
"The most successful AI innovations come from moderate AI performance that’s good enough to provide real value."
"Technology is not inherently bad; many errors come from uneven training and inconsistent configuration that can be fixed."
"When you put a worker at the center and empower them in data collection and product design, the outcome benefits the worker."
"We really need new methods to help designers understand AI as a design material, much like clay or pixels."
"AI innovation has a very small margin of success—80% of AI products fail before market, and 40% of those that launch never succeed."
Or choose a question:
More Videos
"You want to catch accessibility defects early through automated testing in your CI/CD pipeline, what we call shifting left."
Sheri Byrne-HaberAccessibility at Scale
June 9, 2021
"If you skip iterative testing, you risk building features nobody really wants."
Prayag Narula Hannah HudsonEmpowering Designers to do Good Research
March 11, 2022
"Companies had to relearn their customers during COVID because behavior had fundamentally changed."
Janelle EstesUX Research Trends
January 28, 2021
"A single line of UI text might be a small ask but can have massive impact on customer understanding and experience."
Craig Brookes Andreas Huebner Morgan Quinn"Just Make it Look Good" and Other Ways We're Misunderstood
June 11, 2021
"High quality personal connections mean you can engage on a level playing field with serious professionals."
Marc Fonteijn Ru ButlerIncrease your confidence, influence, and impact (through a Professional Community)
December 3, 2024
"The longer you take to recognize research ops and scale your team, the harder it becomes to amplify the value of your research investment — Carrie, Booking.com."
Kate TowseyThe State of ResearchOps: More Than Just Theory
June 20, 2019
"Psychological safety is the number one factor determining team effectiveness and performance."
Alla WeinbergDesign Teams Need Psychological Safety: Here’s How to Create It
September 9, 2022
"The collaboration cycle requires following standards for accessibility, content, design system usage, and IA or you don’t launch your product."
Shawna Hein Kevin HoffmanCreate a Cohesive Civic Design Practice Across Agency, Vendors, and Contracts
November 17, 2022
"Positive mutual regard helps uncover root causes behind people’s behaviors to build collaboration rather than combat."
Tricia WangSCALE: Discussion
June 15, 2018