Is the statement "Claims demand evidence" always true?

[Mudhammam]

No.  I mean that I cannot determine whether any physical system, human or otherwise, has the "property" of qualia, since this property is not measurable or in any way observable.  In the case of other humans, I make what I feel is a pragmatic assumption: they probably experience qualia, too, and more or less than I do.

In order to "observe" qualia, we have to do something like this:
1)  Assume (pragmatically) that other people also experience qualia.
2)  Watch their brains to see what states map to what descriptions of experience XYZ.
3)  Look for the same brain states in others, and ask them if they are also experiencing XYZ.

The problem with this is that even if the assumption in (1) is true, you can't confidently map it to non-human physical systems.  You can't make a computer, claim that it matches human brain states sufficiently, and then claim that your Benotron 2000 experiences what things are like.

Ah, I see. You are referring to the "Hard Problem" of consciousness.  From a first person account of phenomenology, I'm inclined to view it as analogous to the concept of a "nation."  We might say, "I cannot determine whether any government, citizenry, or territory has the "property" of nationhood, since this property is not measurable or in any way observable," in which case "qualia" can be redescribed -- or reduced -- to its more fundamental elements, even if it remains distinct from these, in the same way that a "nation" is neither a government, the citizens, nor a territory, but can apply to all three.  "Qualia", on this view, coincides with 2) and 3) of your suggested "observations."

In terms of the "problem of other minds" and their experience of qualia -- if such is the case -- I concede that it may in principle be impossible to know that others are having similar experiences, as this would amount to knowledge of the subjective side about a particular being from the subjective side of that being; in other words, it requires that it is one's self which is doing the analysis of one's self.  But when it comes to the development of strong A.I., I'm convinced that, though it may always remain a Hard Problem, at a certain point it will simply cease to be an interesting, or even a relevant, one.  This is already true with respects to other human beings, and it is justifiably true in the case of other (higher) animals, which exhibit all of the telltale signs of phenomenology; and we should only expect this, I think, from a biological perspective -- unless there is a Further Fact such as Cartesian Egos which exclusively accounts for our qualia, or, if the human brain (or your brain for that matter) is something awesome that has reached a threshold in its psychophysical evolution that other species (or people) have not attained.  But I don't think it is necessary that all facts be given further justification, which would lead to an infinite regress, and I see nothing erroneous with admitting that our best and most complete description might have to begin with assumptions (though if we apply the "nation analogy" to "qualia" this may or may not be true) -- so long as they remain the most coherent and parsimonious among those available.