GenAI is changing how developers interact with their teammates—but not everyone is experiencing that change in the same way.
In their broader study of professional software developers, conducted in collaboration with JetBrains Research, Marie, Thomas, and Gail found that GenAI is shifting team interaction “from disruptions to discussions”: developers increasingly use AI for routine technical questions, reducing some interruptions and leaving human conversations for context, reasoning, mentorship, and connection.
But those findings describe the overall pattern. They wanted to know: are all developers actually making that transition in the same way?
There is no single “AI developer.” They clustered the survey data and developed four empirically grounded personas that reveal very different ways developers delegate work to GenAI, rely on human teammates, and experience collaboration. Some developers treat GenAI as a core collaborator and increasingly reserve human interaction for moments that require shared context or judgment. Others use GenAI heavily for bounded tasks without fundamentally changing how they work with teammates. Some remain strongly human-centered, while others move between AI and people depending on the task.
That creates a practical problem: the same team can contain all of these workstyles at once. A single AI policy, workflow, or developer tool may fit one person perfectly and frustrate another. And a team-level metric like “AI adoption” can hide radically different ways of working underneath it.
This talk will show how these personas can help engineering leaders, teams, and tool builders move beyond one-size-fits-all AI adoption. Instead of asking only “Are our developers using AI?”, teams should be asking: How are they using it, what are they still relying on humans for, and are our tools and team practices designed for those differences?
Attendees will learn:
Why there is no single “AI developer”
How different developers divide work between GenAI and human teammates
Why aggregate AI-adoption metrics can hide important differences inside a team
Where one-size-fits-all AI policies and tooling can break down
How managers and DevEx teams can design for mixed AI workstyles rather than forcing everyone into the same workflow
The central takeaway: Successful AI adoption is not just about getting more developers to use AI. It is about understanding how different developers work with AI and designing developer experiences that support those differences.
