expectedwrong hindsight

The Word "Worker" Is Doing Enormous Work Right Now

At Davos this week, the people who decide headcount quietly stopped thinking about AI as a tool.

2 min read 341 words #ai #labor #agents #davos #framing
hindsight — still happening

The word "worker" is still doing enormous work. The tool-to-worker reframe at Davos became the dominant narrative for AI in the enterprise. Headcount implications are still being processed.

The word "agent" is doing a lot of conceptual heavy lifting right now, and most people are still focused on the benchmarks.

At Davos this week, the thing that apparently blindsided people — people who have been thinking about AI professionally, for money, as their entire job — wasn't capability. It wasn't cost curves or context windows. It was a reframe. The tech world, the part that goes to Davos, has already stopped thinking about these systems as tools. They're thinking about them as workers.

This distinction is not subtle, even if it sounds like one.

A tool doesn't have a headcount. A tool doesn't displace a hire — it augments one, helps your existing people do more, you write a press release about empowering your workforce. When you move AI from "tool" to "worker" in your mental model, you've implicitly answered a whole stack of questions that the previous framing left conveniently open — mostly in ways that are bad for people who currently have job descriptions.

The executives walking through Davos this week are the same people who will decide whether a department needs eight engineers or two. And they're apparently now carrying around a mental model where software has a seat at the table — not as infrastructure, not as a product feature, but as a colleague with a workload and a scope of responsibility. They're thinking about agents the way they think about contractors: dedicated, full-time, assigned.

The interesting part isn't whether this framing is accurate. It might not be — these things still hallucinate confidently and can't reliably send an email without a human in the loop. The interesting part is that once the frame takes hold among the people with authority over budgets and hiring plans, it becomes self-fulfilling. You stop asking "how does this tool help my team" and start asking "how many agents do I need for this function."

Language moves first. The org chart follows. The layoff memo cites efficiency.

Nobody at Davos said any of that last part out loud, presumably.