One of the most visible problems with AI discourse is a vocabulary problem.
We are not only using AI; we are continuously reclassifying it mid-interaction. Is this thing helping me, manipulating me, reflecting me, completing me, replacing someone, pretending to understand, acting as software, or acting as a social presence?
A large amount of cognitive effort is spent silently translating. We are often not disagreeing only about AI itself; we are disagreeing through different borrowed vocabularies. We call AI a tool, an assistant, an agent, a partner, a machine, a collaborator, an oracle, a mirror, a stochastic parrot. None fit perfectly, each word captures something and each word distorts something. Each imports a hidden moral and practical script.
We have had partial language for non-human agency for a long time: spirits, animals, markets, bureaucracies, institutions. But we have very little ordinary language for highly capable conversational systems that participate in cognition without being biological minds.
In some ways, this is similar to an older conceptual problem. For a long time, people could see trees, animals, rivers, and soil. But “ecosystem” required a different kind of perception: a relational one. Once the concept existed, things that had always been present became easier to see. The forest stopped being a collection of objects. It became a network of interactions.
AI may be creating a similar conceptual gap. The comparison is not that AI is like nature, but that both require concepts that make relations visible. We can see the interface, the answer, the prompt, the model. But we still lack ordinary language for the relation itself: a non-human conversational system that participates in cognition; a tool that behaves socially; an infrastructure that enters thought at the level of language.
So maybe the problem is not that AI is hard to name, but that it creates a relation we do not yet know how to see. From this point of view, the question becomes: what kinds of relations do these systems make possible, and what words would allow us to see them clearly?