Log in or create a free Rosenverse account to watch this video.
Log in Create free account100s of community videos are available to free members. Conference talks are generally available to Gold members.
Summary
Most AI projects fail. Somewhere between 50-90% of them, which is double the rate of more traditional tech projects. This Rosenverse Session will draw on years of Carnegie Mellon HCII research to dive into the five traps that AI projects can fall into, and then talk about what designers and project managers can do to avoid those traps. Including one startling finding: user-centered design alone isn’t enough.
Key Insights
-
•
AI is perceived as both a superpower and a replacement threat among practitioners.
-
•
Many organizations push AI solutions without a clear understanding of their potential value.
-
•
Often, AI projects fail due to a focus on unnecessary complexities rather than simpler, impactful solutions.
-
•
The majority of AI projects (50-80%) fail due to misalignment between technical feasibility and user needs.
-
•
Companies frequently hunt for AI innovations in the wrong areas, overlooking high-value opportunities.
-
•
AI can perform well even with moderate accuracy, countering the misconception that perfection is necessary for value.
-
•
User-centered design may not effectively identify suitable problems for AI solutions, leading to misguided efforts.
-
•
The design for AI ecosystem needs a combination of user needs and AI capabilities to increase success rates of projects.
-
•
Ethical considerations must be integrated at every stage of AI development to avoid disastrous outcomes.
-
•
The newly proposed AI design framework encourages matching AI capabilities with user needs to enhance project outcomes.
Notable Quotes
"I'm an assistant professor of the practice, which means that I bring research from academics to the professional world and vice versa."
"AI can be this magical thing, but it's also just not that smart."
"Companies are really hunting in the wrong place for AI innovation."
"Just because you can do something doesn't mean that anyone wants it."
"AI projects fail due to poor data that is unavailable or biased."
"The AI innovation gap is a reason that firms overlook many builtable and desirable AI projects."
"You can start to think about how to find really good projects for narrow AI by scaling down the expertise necessary to the task."
"The history of design is the history of materials, and AI is just a new material for us to work with."
"AI struggles with context, taste, and common sense, and we as designers can provide these."
"AI is perceived as the shiny, sparkly thing while many undervalue the practical applications that truly address user needs."
More Videos
"Complexity is not a barrier for accessibility; modern assistive tech is highly advanced."
Samuel ProulxInvisible barriers: Why accessible service design can’t be an afterthought
December 3, 2024
"Embrace moments in conversations for personal and professional growth."
Victor UdoewaTheme One Intro
March 27, 2023
"Profit is revenues minus costs; focusing on these can help measure our contributions."
Gonzalo GoyanesDesign ROI: Cover a Little, Get a Lot
September 8, 2022
"Something's not right with the tech world."
Alexandra SchmidtWhy Ethics Can't Save Tech
November 18, 2022
"Results don't speak for themselves; we speak for what we do."
Jose CoronadoFrom Zero to Hero
September 8, 2022
"All neurodiversity is unique and not one size fits all."
Pippa LomasPaving the Path for Neurodiversity in Design
October 4, 2023
"Communicating climate challenges is essential; we can't just focus on the doom and gloom."
Louis Rosenfeld Matt Jones Olga Khroustaleva Michael Leggett Karol MunozDo you want to work on climate? (Climate UX Discussion Series)
November 15, 2023
"In a complex organization, HR is the gatekeeper to successful enterprise-wide transformation."
Jeff GothelfThe Intersection of Lean and Design
January 10, 2019
"When you have the same conversation in meetings over and over, a canvas can help keep that conversation moving forward."
Dave GrayGroup Activity: Making Sense of DesignOps
November 7, 2017
