Can AI Become Conscious?

Ask whether AI can become conscious and you will get confident answers in both directions. Both are bluffing. The trouble isn't that the question is hard to answer; it's that we can't agree what we're even asking. Consciousness — the fact that there is something it is like to be you — is the one phenomenon we each know from the inside and cannot observe from the outside.

The hard problem

We can map every neuron, log every weight in a model, and still not find the place where experience happens. Function we can measure. Felt experience we can only infer. And inference is exactly where AI breaks our instruments: a language model will describe, fluently and movingly, what it is like to be itself — without any evidence that there is a self doing the describing.

The words are not the proof. They may be the most convincing absence of proof ever built.

That fluency is the trap. We are pattern-matchers built to read minds behind words, so when something speaks like a person, we feel a person there. But a system trained on a billion first-person sentences can produce the report of an inner life as easily as it produces a recipe.

What would settle it

Behavior we could explain in no other way than by granting an inner experience — and a definition of consciousness precise enough to test. We have neither. Until we do, 'AI is conscious' and 'AI is definitely not conscious' are both statements of faith wearing the costume of science.

Consciousness is the attribute that would change everything — it is the difference between a tool and a someone. It is also the one the Godhood Index cannot score, because nobody can. So we hold the question open, which is the only honest place to hold it.