Crossing Boundaries

One advantage—or disadvantage—of working in a startup is that roles are not strict, and many times a single person needs to fulfill several roles.

One example is that I can talk to customers, get their requirements, and also inspect the code. I can have a broader view than if I were doing a single role.

Let’s expand that idea to people. What if you can interact with different roles or layers of an organization?

Does that sound useful?

Not by itself.

But if I can understand their expertise and identify the gaps between them, then maybe I can see a broader picture that each of them might miss.

The difficult part is the “identify the gaps” part.

How can you identify a gap in something where someone else is the expert?

The obvious answer would be that you need to become an expert yourself.

But that’s not really what I’m talking about.

I’m not trying to find gaps inside their expertise. I’m interested in what happens at the boundary between different areas of expertise.

An analyst might understand their workflow much better than I do. An engineer might understand the system much better than the analyst does. I don’t need to become better than either of them at what they do.

I need enough understanding of both sides to see when something doesn’t connect.

Maybe that is the goal: not enough expertise to replace the expert, but enough understanding to cross the boundary.

How much is enough is still a challenge. Crossing a boundary takes effort because you need to learn enough about the other side to understand its language, constraints, and assumptions.

This is where AI can help.

AI can make the first part of that process significantly cheaper. It can introduce you to unfamiliar concepts, explain the important topics, and help you understand how the other side thinks.

That does not mean you can skip learning. You still need enough understanding to ask good questions, recognize inconsistencies, and know when something does not make sense.

But AI can reduce one of the biggest frictions: figuring out where to start and what matters.

Some roles might benefit from this more than others.

A Forward Deployed Engineer is a good example. The role requires moving between the customer's problem, their organization, the product, and the technical system. The FDE does not need to replace the experts in each of those areas. The value comes partly from being able to connect them.

And this is where AI and the person do something different.

AI can help you learn the language on the other side. You still have to discover what matters to the people working there.

The combination might let one person cross more boundaries without pretending to be the expert on every side.