Faster alone, further together: Rebuilding collaboration in the age of AI research
Summary
As AI agents accelerate early-stage product work—research synthesis, concept generation, and opportunity analysis—teams risk losing what makes research effective: shared understanding. Individuals can now produce artifacts that once took days of team discussion and critique, resulting in faster outputs but weaker collective insight. On LinkedIn’s Growth org, we saw this tension as teams adopted AI-powered research, design, and strategy tools. To address it, the UX Research team built a JTBD-based Competitive Analysis workshop paired with an AI teammate—not to replace collaboration, but to scaffold it. This pair of tools structures how teams jointly explore 0-to-1 opportunities, align on member Jobs-to-Be-Done, and analyze competitive differentiation with an AI agent in the loop. This session shares a leadership perspective on re-engineering collaboration around AI rather than letting AI erode it. I’ll show how structured, AI-assisted workshops enable both rigor and creativity—and how these practices have elevated the pace, quality, and cohesion of strategic decision-making across teams.
Key Insights
-
•
Growth PMs and UX researchers operate from fundamentally different theories of evidence, causing misalignment in priorities.
-
•
Organizational systems are optimized around PM processes, not UX research frames, privileging metrics over human-centered insights.
-
•
Common growth strategies like doubling notifications or copying features from other apps stem from short-term metric focus, not from deep problem understanding.
-
•
Simply translating research insights into PM language can lose the valuable context and reframing UX research provides.
-
•
AI tools can help bridge the gap by reasoning across both growth metrics and human-centered paradigms simultaneously.
-
•
Trust and shared understanding develop from cooperative processes, not just from clearer communication or presentations.
-
•
Structuring workshops with individual ideation, AI-supported refinement, and group negotiation helps build shared vocabulary and alignment.
-
•
Designing collaborative reasoning conditions is a key role for UX researchers beyond traditional insight delivery.
-
•
Clear agenda design, curated inputs, and defined roles improve the efficiency and outcomes of cross-paradigm workshops.
-
•
AI can act as a cognitive accelerant, providing healthy constraints and expanding thinking, fostering creativity in teams.
Notable Quotes
"This is a story about what happens when you stop trying to will others to think like you and start creating conditions under which both frames can coexist productively."
"Growth PMs think this way because it’s the only kind of thinking the system we’re working within can operationalize."
"The core question UX researchers ask is did we solve the right problem for the user, while PMs ask did the metric move."
"When you boil down all the research into actions someone might take, the thing that made our research valuable in the first place, understanding context, was all gone."
"No amount of clearer presentation or communications is going to solve this; it’s a cooperation problem."
"We’re not just translators between different ways of thinking, but architects of shared reasoning processes."
"Giving people time and space to think individually and pressure test with AI before group discussion prevents dominance by power."
"Sometimes our most valuable work isn’t the insights we produce, but the conditions we create for shared reasoning across paradigms."
"AI can provide healthy constraints, help people focus, use a common language, and act as cognitive accelerants."
"The real breakthrough was when the pilgrim’s mechanical bird spoke both ways, braiding their words together to reach the summit."
Or choose a question:
More Videos
"What if everyone was accepted? We could rise almost anyone to proficiency if the curriculum and pedagogy are strong enough."
Nathan Shedroff Hugh Dubberly Thomas J. McLeishHow Will Design be Taught When the Schools Shut Down?
May 8, 2026
"Tracking and capturing the work can be a whole new work in itself. People will naturally abandon complicated process."
John Paul de Guzman10k Screens Later: How We Became a Data-Driven Design Organization
September 24, 2024
"Advanced logic, matrix questioning, AI survey design creation — these are baseline capabilities you should expect from any modern survey tool."
Aleksandra Korczynska Caroline Jarrett Justyna ParmeeSurvey Tools
March 12, 2026
"Motivation and focus are especially challenging when working from afar."
Kristin SkinnerTheme 2: Introduction and Provocation
January 8, 2024
"Attendee cohorts meet daily in small groups with facilitators to share learning goals and deepen engagement."
Rachael Dietkus, LCSW Victor Udoewa Jennifer StricklandEverything You Need to Know about the Civic Design 2022 Call for Presentations
May 17, 2022
"Are we feeling trapped in this thing called UX design, limited by mechanized processes like agile?"
Uday GajendarThe Rise of Meta-Design: A Starter Playbook
May 19, 2022
"The hidden hero or villain of craft is really feedback, the quality of feedback that we give."
Vanessa VarinFeedback: The Other F-Word
September 10, 2025
"Horst Rittel introduced wicked problems to describe problems where consensus on definition is difficult, and calling them problems can be misleading—maybe they’re better called messes or tangles."
Hugh DubberlyProblems with Problems: Reconsidering the Frame of Designing as Problem-Solving
June 19, 2019
"The entire customer experience was broken, users were frustrated, support teams were overwhelmed, and internally teams were struggling to align."
Sean McKayCoexisting with non-researchers: Practical strategies for a democratized research future
March 11, 2025
