I think saying "subjective experience is private" is just expressing an analytic truth.
I'm not sure this is right, actually. Consider a least convenient case: a world populated by conscious beings (such as AI's) whose subjective experience is actually made up of simple numbers, e.g bytes stored in a memory address space. (Of course this assumes that Platonic numbers actually exist, if only as perceived by the AI's. Let's just concede this for the sake of argument.) Suppose further that any AI can read every other AI's memory. Then the AI's could know everything there is to know about each other's experiences, yet any one experience is still "subjective" in a sense, because it is associated with a single individual.
But such experiences still aren't subjective in the sense of "private". I don't see what you are getting at. If subjective=private, your AIs don't have subjective experience. Setting up another definition of subjective doesn't stop subjective=private from being analytically true or true at all. There are lots of things associated with individauls, such as names, which are not subjective.
[Cross-posted.]
1. Defining the problem: The inverted spectrum
A. Attempted solutions to the inverted spectrum.
B. The “substitution bias” of solving the “easy problem of consciousness” instead of the “hard problem.”
2. The false intuition of direct awareness
A. Our sense that the existence of raw experience is self-evident doesn’t show that it is true.
B. Experience can’t reveal the error in the intuition that raw experience exists.
C. We can’t capture the ineffable core of raw experience with language because there’s really nothing there.
D. We believe raw experience exists without detecting it.
3. The conceptual economy of qualia nihilism pays off in philosophical progress
4. Relying on the brute force of an intuition is rationally specious.
Against these considerations, the only argument for retaining raw experience in our ontology is the sheer strength of everyone’s belief in its existence. How much weight should we attach to a strong belief whose validity we can't check? None. Beliefs ordinarily earn a presumption of truth from the absence of empirical challenge, but when empirical challenge is impossible in principle, the belief deserves no confidence.