Working with AI Feels More Like Leadership Than Coding

There's a delicious irony in the idea that the most cutting-edge act of programming is beginning to resemble the most ancient human skill: leadership. For decades, we imagined that working with machines meant speaking their language—rigorous, deterministic, and utterly obedient. But large language models have broken that contract. They don't execute; they respond. They don't follow instructions; they interpret them. And suddenly, the human at the keyboard is less of a coder and more of a captain steering a ship with a mind of its own. This reframing, articulated in the essay "Working with AI Feels More Like Leadership Than Coding" (https://allen.bargi.org/notes/working-with-ai-feels-like-leadership/), touches something profound about the nature of agency. When you delegate a task to a model, you're not just offloading computation—you're entering into a relationship with a statistical echo of human intent. The outputs are probabilistic, shaped by billions of parameters we can't fully trace. To get good results, you don't write a spec; you paint a vision. You clarify values. You give feedback on vibes. In other words, you lead. This feels weird because it blurs a line we thought was solid: the line between tool and collaborator. A hammer doesn't suggest a better way to hit a nail. But an AI might offer a different approach, challenge your assumptions, or fail in ways that force you to question your own reasoning. That's not a tool; that's a sparring partner. It forces us to confront the possibility that intelligence—even synthetic intelligence—is not a switch we flip, but a spectrum we participate in. We are no longer just observers of a system; we are co-creators of a process that has its own emergent behavior. And here's where the wild part kicks in: if working with AI is leadership, then what does that say about leadership itself? Perhaps the skills we're developing—patience, clarity, empathy for a non-human mind, and the ability to hold ambiguity—are the same skills we need to navigate a universe that is itself probabilistic and strange. Quantum mechanics taught us that observation shapes reality. Now, our daily work is teaching us that intention shapes intelligence. The essay is a small note in a vast conversation, but it points to a seismic shift in how we define our place in the cosmos: not as the sole thinkers, but as the shepherds of synthetic minds, guiding forces we barely understand toward outcomes we hope to own.