The case for hybrid teams

I've spent a lot of time lately with a question that sounds simple: what is an AI agent? Is it a tool, another piece of software, a utility? None of those descriptions quite captures how I engage with AI every day in my professional life.

The agents I work with have jobs. We communicate with them, and they communicate with us. They are thought partners, subject matter specialists, and doers. I found myself looking for a way to describe that working relationship, and colleague felt right, even if it felt a little awkward at first.

A colleague has a place in the work alongside us. We depend on their contribution, give them context, and figure out how our responsibilities fit together. Thinking about agents this way opens up the possibility of a hybrid team: people and AI agents contributing to a shared outcome.

Here, hybrid describes the mix of human and AI contributors, rather than where people happen to work. People remain responsible for the team's direction and decisions. The agent brings no human experience to that work, but we can give it a defined role within it.

Quinn, my AI Chief of Staff, gives me a practical starting point. I built Quinn with organizational context and safeguards to help me focus on the work that matters. The role helps me judge its usefulness: whether it helps me give more attention to my team and our strategy.

Across a team, that contribution might look like an agent helping prepare for a project meeting by gathering updates and identifying unresolved questions. With that preparation checked and available, people could spend their time working through a disagreement or deciding what needs to change. The preparation earns its value through what it enables the people to do together.

Thinking of agents as colleagues encourages us to design for that relationship. A team can have individually capable members and still struggle because their work doesn't fit together. An agent's output can be technically correct and still leave someone with more work to untangle. Giving it a place in the team means considering who depends on its work and whether the contribution is actually helping them.

That is what interests me about hybrid teams. I want us to build working relationships with AI that make the day easier to navigate and leave people with more attention for each other. Getting there starts with a familiar leadership responsibility: thinking carefully about the team we are building and what we want it to accomplish.