Working with AI Is More Like Leadership Than Coding

In a personal essay published August 15, 2026, Allen Bargi contends that traditional software gave engineers certainty: the same input always produced the same output, and deviation was a bug. AI, by contrast, is non-deterministic — the same prompt can yield different results, surface an unexpected approach, or miss an obvious point. Bargi draws a parallel to managing people: a colleague might deliver exactly what was asked, or something better because they grasped the underlying intent. He argues that treating AI like a compiler leads to frustration, while treating it as a collaborative interaction improves outcomes. He is careful to note the comparison is about working style, not personhood — AI lacks lived experience, accountability, and human judgment. The practical implications he outlines mirror leadership best practices: share context, explain desired outcomes, set boundaries, and respond to what comes back. Beyond single prompts, Bargi emphasizes the value of a shared working context — examples, corrections, and reusable instructions that align the system with the user's thinking over time. He frames the broader shift as moving from issuing commands to expressing intent, noting: 'We spent years learning how to tell computers exactly what to do. Now we also need to explain why the work matters.' He concludes that the technology is new, but the underlying leadership skills are not.