The one human employee

About

I started as a writer. I wanted to make the whole thing. Now I do.

The arc

I've always had an entrepreneur's streak: the instinct to build the thing that doesn't exist yet instead of picking one off the shelf. Cornell caught it early. Freshman year I was invited to join the College Scholar Program, where a handful of students design their own liberal-arts major from scratch. One requirement: a senior thesis that ties it all together.

I interned at MTV in college, then spent the next stretch of my career in the digital departments of major media brands: NBC, the NHL, Disney. I came in at NBC as a writer in the News department, and the work grew more hands-on from there: at the NHL I ran eight team sites while filing three features a week for NHL.com, and by Disney I was building a consumer product that outside analysts rated the easiest to use.

What kept pulling at me was how things actually got built and shipped. So I started my own agency, Bandwidth Productions, where I've owned the specs, the plans, and the accountability while a team of developers and designers did the building. Our clients were a mix of institutions, media, celebrities, and young artists: VH1, the Cathedral of St. John the Divine, the Museum of the Moving Image, the Scholastic Art & Writing Awards, Bobby Flay, and Giada De Laurentiis. On some projects we collaborated with the design legends of Pentagram. The editorial writing didn't last, but the skill underneath it did: saying exactly what you mean. Nothing I've learned since has mattered more. Long before I built a fleet of AI agents, I'd already learned the lesson I still rely on: good work comes from a clear idea, a precise brief, and people with excellent judgment.

When AI matured into something you manage like a team rather than operate like a tool, everything I'd practiced for years suddenly compounded. I stopped being the person who needed designers and developers and became a person who ships: a live product for income investors, and a growing collection of systems that run my information diet, my finances, and my family's travel.

Now

I spend enough time around technology. I'm usually more interested in people who make something entirely different, and in how they think.

Once a month, I host a session on AI for Long Angle, a private community of founders and investors, usually with 50 to 100 people on the call. Seeing what everyone in the group is building, and getting to share my own, is one of the more remarkable hours of my month.

The thesis

For a while, computer science was the degree to have. Then the machines learned to write the code, and the skills that kept their value were the liberal-arts ones: reading a problem, saying exactly what you mean, knowing what's worth making at all. The major I designed for myself at eighteen turned out to be a better bet than I knew.

Here's where I think this is going, starting now and accelerating: the question stops being what you, individually, can contribute to a project or a company. It becomes what your team can contribute. You, plus the team you bring with you.

That's what my orchestration actually is. I'm actively building that team and strengthening every component of it. Counter to the popular assumption that better models will make orchestration frameworks obsolete, I believe the opposite. Every time the foundation models improve, my team gets stronger alongside them. The infrastructure isn't replaced by better models; it's amplified by them.

So I intend to keep building it. I have a team with zero turnover that only grows in experience and in size. I'm more curious than I've ever been about what comes next.

Psst, what is Humans Only?Look for me? →