Like some other people seem to do, I mostly seem to tend to use the
labels "conscious" and "self-consciousness" for different (but each only
vaguely defined) things. Self-consciousness (as a vague, not
well-defined concept) is regarded by me as a more "complex" thing than
"consciousness" (also a vague, not well-defined concept).
Thinking further however, I seriously wonder whether this definition-of-terms stuff -- as long as these definitions ALL remain vague -- really solves anything.
To put in my favored thermostat example again:
In my opinion one would have to agree that it is plausible to regard even a thermostat as "conscious" : The dynamic operation of the thing is such that it's "aware" of the temperature around it in relation to its "goal" of keeping the temperature near its "target" temperature. That is: if one REGARDS the thermostat as a "living, conscious thing", then it's in my opinion easy to exactly point out this conscious behavour of the thermostat, and even to identify in the thermostat's operation things like "goals" that in HUMANS we also tend to speak about when we talk about human consciousness.
That is: a large part of the difficulty with this consciousness stuff seems to reside in "ourselves", i.e. in what we -- w.r.t. our own beliefs -- are prepared and inclined to regard as "conscious".
An analogy : 300 years ago, white Europeans seem to have not regarded black negroes as very "conscious". What people regard as "conscious" depends often very much not on the entities considered, but on the "beliefs" in the heads of those who do the considering. I think we need to be very aware of this.
I'm wondering whether with the same right as a thermostat can be called "conscious", it couldn't also even be called "self-conscious". It seems easy to me now to identify even a "sense of self" in the thermostat (although I agree that this is pulling the analogy further than with the "simple" consciousness). The "sense of self" of the thermostat is how it "sees" itself in relation with its environment. The thermostat's "sense of self" could be said to consist in "being an entity that aims to keep the temperature near temperature X". If one, e.g., would change the temperature setting of the thermostat, one 'd change the thing's "sense of self".
So since (in my opinion) it can be stretched this far what one can "plausibly" regard as (self-)conscious, this (to me) demonstrates that the main difficulty with (self-)consciousness could very well reside in our own beliefs on them, instead of in the machines/entities/mechanisms/animals/beings that we study and assess the consciousness of, and/or that we aim to create in order to create "conscious AIs".
In view of the above, I therefore also think that the question is not trivial as to whether maybe effectively almost ALL of consciousness or self-consciousness could reside in these "you see what you want to see" interpretation effects in our own minds as the people who are analyzing/assessing/considering an entity and judge how (self-)consciousness we find it.
That is: I think we have to ask ourselves if it could be the case that it is to the full 100% a matter of "the emperor's clothes", i.e. matter of interpretation, and not a matter of a technical machine to be constructed. Because we can already regard very simple machines that already exist as (self-)conscious when only we "stretch" our concepts of (self-)consciousness enough. To repeat: I think that engaging in such "stretching" of these concepts is NOT clearly a case of "playing false", BECAUSE of the very fact that there seem to exist no clear-cut definitions of these concepts. (The very "vagueness" of the concepts, I think, straightforwardly invites or even necessitates stretching them : to define means to discover the limits hence to stretch till one discovers a clear-cut limit. If one discovers no such clear-cut limits, maybe one has to admit that none exist.)