Turing’s Cat

Schrödinger × Turing

An entity that stays in superposition: neither conscious nor blind, neither tool nor interlocutor.

It is intelligent when it writes moving poetry. It is stupid when it puts six fingers on a hand. It is brilliant when it solves math problems. It is idiotic when it cannot count the R’s in strawberry.

It is not that it is sometimes intelligent and sometimes stupid. It is that it exists perpetually in both states at once. Measurement, the test, the use, the judgment, collapses it temporarily, but does not resolve the ontological question.

And perhaps that is right. Perhaps we need not resolve whether AI is intelligent or not. Perhaps we must learn to work with an entity that stays in superposition, one that can be useful even without being “intelligent” or “understanding” in our sense.

Turing’s Cat is AI as an entity in permanent epistemological superposition. That impossible creature born from a friend’s mistake, who confused Schrödinger with Turing, the quantum cat with the imitation game. A mistake that turned out to be a lens.

The gesture: do not collapse immediately to the answer. We pause in useful perplexity, where you can see what changes in our own behavior when we speak with a model. This is the minimal ethics the Cat suggests: recognize the doubt, tell it, decide with the body.

When the AI’s output arrives, there is always the temptation to close the doubt: “does it work or not?” The Cat teaches that the question is ill-posed. It works in certain contexts, fails in others, and the boundary is mobile, not binary. The Confident Attitude is born here: from accepting to operate in a regime of superposition without the anxiety of resolving it.

Inhabit the superposition, do not resolve it. Intelligence as relation, not as substance.

→ Full treatment in Chapter 5: Turing’s Cat.