Coexisting with AI: A practical guide for researchers to navigate tools, ethics, and integration
This video is featured in the PDI 2026 playlist.
Summary
Key Insights
-
•
AI is shifting the authoritative voice in research from humans to generative models, challenging researchers’ traditional value.
-
•
Framing research value by sheer output risks commoditizing the field and losing long-term business impact.
-
•
Measuring research impact through concrete business outcomes is harder but essential for sustaining researcher relevance.
-
•
Research integrity is at risk if AI adoption isn't rigorously tested, as stakeholders expect researchers to model best practices.
-
•
AI tools can introduce biases especially affecting marginalized groups, impacting fairness in user research.
-
•
User researchers’ specialized skills in navigating complex trade-offs uniquely position them to integrate AI responsibly.
-
•
The lack of standardized methods for using AI in social science research makes learning from engineers and scientists critical.
-
•
Human data collection remains irreplaceable due to AI's weaknesses in understanding niche cultures and novel insights.
-
•
AI enables researchers to extend beyond study execution into prototyping, impact tracing, and direct product decision influence.
-
•
The rise of AI-powered 'triple threat' researchers combines research, product management, and design augmented by AI capabilities.
Notable Quotes
"Right now, there’s a fear of being replaced by AI, and while a lot of that’s overblown, researchers have a legitimate reason to feel uncomfortable."
"The choices we make about AI are about much more than speed versus quality — they define how we value research and our relationships."
"Volume is a losing game because we cannot outcompete full AI automation for sheer speed and quantity of insights."
"If we want our value to be defined as anything other than studies run, we must learn how to trace and articulate the actual business impact."
"If we are haphazard about AI adoption and trust marketers over rigorous testing, it looks like all our hand wringing about research accuracy was just gatekeeping."
"Generative AI is non-deterministic and doesn’t understand meaning, so baseline error rates exist in all tools we use."
"Prompting is a specialized skill, just like statistics or in-depth interviewing; the wrong wording can bias your results."
"More than a third of content on social media sites is AI generated, creating a training problem for new AI models."
"AI can empower researchers to solve pain points — like anticipating trends or critiquing discussion guides — amplifying impact rather than replacing work."
"We should think of AI as something that amplifies what we already are: if we think research is about impact, AI will help us understand and track it."
Or choose a question:
More Videos
"Space has meaning in a map — moving components changes the map's meaning, unlike diagrams where space is arbitrary."
Simon WardleyMaps and Topographical Intelligence
January 31, 2019
"Visualizing carbon emissions on user journeys provokes conversations that numbers alone might not evoke."
James ChudleyDecarbonising User Journeys: How minimising enables us to do more with less
February 19, 2025
"What if we had AI teammates who were on every call and could help moderate, recap, and track time? That future is likely."
Bria Alexander Laura Gatewood Corey Long Daniel Orbach Laine Prokay Deanna SmithThe Big Question about Resilience: A panel discussion
September 23, 2024
"Residents show up with their full selves, wanting solutions for multiple needs, not just the narrow project scope."
Ashley Cortez Devika Menon Nidhi Singh Rathore Danita J. Reese Mari NakanoShifting Toward Community-Led Innovation in Local Government
December 8, 2021
"Design Operations is a massive force multiplier on your design team—it can increase output by at least 25 to 30 percent."
Bob BaxleyLeading with Design Operations Past and Present
December 19, 2019
"The time for talk has passed; we need immediate action and collaboration."
Alex Hurworth Bonnie John Fahd Arshad Antoine MarinDesigning a Contact Tracing App for Universal Access
October 23, 2020
"An SOW at LinkedIn takes six to eight weeks, and by then sometimes the project isn’t even relevant anymore."
Jon Temple Kalee Dankner Bruce Falk Lauren GalanterPanel: Stacks, Security, and Stakeholders: The Hidden Work of UXR Tool Procurement
March 12, 2026
"We’re not trying to change behavior, but help researchers in their existing processes to make research faster."
Kate Towsey Basel Fakhoury Oren Friedman Graham GardnerParticipant Recruitment and Management Tools
March 12, 2026
"Ethics evolve faster than laws; just because something is legal doesn’t make it ethical."
Erin WeigelGet Your Whole Team Testing to Design for Impact
July 24, 2024
