‘Management’ has never been, nor will it ever be, about people. It has always only ever been about things.

The term originates as a way of describing methods for controlling a horse. Over time, the term shifted to involve the administration of household resources (remember that our term ‘economics comes from the Greek oikos, meaning home). During the Industrial revolution, the term shifted again in order to describe more scientific approaches to coordinating people (i.e. the activity of human bodies as productive things) in order to achieve well-defined outcomes.

(consider, for a moment, the idea of ‘human resources,’ which does not primarily involve care for the human soul, but rather mobilizing humans as things in order to accomplish economic outcomes).

So, management has always only ever referred to methods for predicting and controlling the behavior and outcomes of things.

Given this, I am not at all bothered when people start talking about the management of AI Agents. Agents are things we aim to control in order to predict and define a limited set of outcomes. The term ‘management’ here is perfectly apt.

What is offensive, then, is not taking a ‘human’ concept like management and applying it to things. Instead, what is offensive is the fact that, since the industrial revolution we have talked about managing people AS IF they were things.

People management is what happens when slaves are not an option, and automation is neither sufficiently powerful nor reliable.

In the true sense of the term, management is a term that can only apply to humans to the extent that they are treated as things.

And yes, you are managed today by the company you work for, which aims to predict and control your behavior and outputs in order to achieve ends that are rarely your own.

Long story long: yes, you are managing agents. And the future will involve managing agents in a way not dissimilar to how slaves were managed in ancient Greece. In fact, I would argue that AI Agents are an attempt to fulfill the promise of Aristotle — who saw slaves as a way to offload responsibilities concerning the maintenance of bare life in order to allow men (women and children were classed similarly to slaves) to contemplate the Good all day long — but in ways that are less — or at least differently — fraught from an ethical perspective.

Is the problem, then, that we are managing agents? Or does the opportunity to manage agents now make it possible for us to stop managing people in ways that — insofar as our aims have been about prediction and control — were never actually particularly humane in the first place?