If anything has, minimally, a robust spatiotemporal sense of itself, and can project that sense forward to evaluate future outcomes, then it has a robust "self."

What this requires is a persistent internal model of: (A) what counts as its own body/actuators/sensors (a maintained self–world boundary), (B) what counts as its history in time (a sense of temporal continuity), and (C) what actions it can take (degrees of freedom, i.e. the future branch space), all of which are continuously used to regulate behavior under genuine epistemic uncertainty. When (C) is robust, abstraction and generalization fall out naturally. This is, in essence, sapience.

By "not trivially reducible," I don't mean "not representable in principle." I mean that, at the system's own operative state/action abstraction, its behavior is not equivalent to executing a fixed policy or static lookup table. It must actually perform predictive modeling and counterfactual evaluation; collapsing it to a reflex table would destroy the very capacities above. (It's true that with an astronomically large table you can "look up" anything -- but that move makes the notion of explanation vacuous.)

Many robots and AIs implement pieces of this pipeline (state estimation, planning, world models,) but current deployed systems generally lack a robust, continuously updated self-model with temporally deep, globally integrated counterfactual control in this sense.

If you want to simplify it a bit, you could just say that you need a robust and bounded spatial-temporal sense, coupled to the ability to generalize from that sense.

The zombie intuition comes from treating qualia as an "add-on" rather than as the internal presentation of a self-model.

"P-zombie" is not a coherent leftover possibility once you fix the full physical structure. If a system has the full self-model (temporal-spatial sense) / world-model / memory binding / counterfactual evaluator / control loop, then that structure is what having experience amounts to (no extra ingredient need be added or subtracted).

I hope I don't later get accused of plagiarizing myself, but let's embark on a thought experiment. Imagine a bitter, toxic alkaloid that does not taste bitter. Suppose ingestion produces no distinctive local sensation at all – no taste, no burn, no nausea. The only "response" is some silent parameter in the nervous system adjusting itself, without crossing the threshold of conscious salience. There are such cases: Damaged nociception, anosmia, people congenitally insensitive to pain. In every such case, genetic fitness is slashed. The organism does not reliably avoid harm.

Now imagine a different design. You are a posthuman entity whose organic surface has been gradually replaced. Instead of a tongue, you carry an in‑line sensor which performs a spectral analysis of whatever you take in. When something toxic is detected, a red symbol flashes in your field of vision: “TOXIC -- DO NOT INGEST.” That visual event is a quale. It has a minimally structured phenomenal character -- colored, localized, bound to alarm -- and it stands in for what once was bitterness.

We can push this further. Instead of a visual alert, perhaps your motor system simply locks your arm; perhaps your global workspace is flooded with a gray, oppressive feeling; perhaps a sharp auditory tone sounds in your private inner ear. Each variant is still a mode of felt response to sensory information. Here's what I'm getting at with this: There is no way for a conscious creature to register and use risky input without some structure of "what it is like" coming along for the ride.