It is not a question of what the sensor detects. It is a question of how it is possible for there to be an experience when sensing occurs.
Your introspection is simply pointing out the likely nature of what you experience, and I actually agree (tentatively) with the idea that most of our conscious experience is rooted in a self-reflecting system. But none of that has any bearing on how there can be ay experience at all.
Think about it: How do you know that you have what you call an “experience”? It’s because you perceive the having of the experience. So, at some point, the perception of having an “experience” is something that enters as an input into your cognitive process, and you match it to some mental models you have about such inputs.
I adjusted my mental model to think of those “qualia” perceptions as being the sensing of parts of the internal workings of my brain. It’s a side-effect of all the processing that is going on, if you will, and of the likely fact that the sensing of some subset of the processing steps is being feed back as inputs into the cognitive processing.
We know that physical systems can sense phenomena in the world. We doubt that they experience anything when they do so. Even if we create a Hofstadterian "strange loop" so that we sense our own sensing, that does not give rise to "experience".
I concede that it could be a matter of scale, but a photosensor that knows that it glows yellow when it senses red does not have an experience.
Yes, a photosensor doesn't have that kind of sensation, and that's because there is nothing going on inside of it that would correspond to that sensation. In the humain brain there is enough that is going on (the complexity and quantity is staggering) that our range of inner experience is easily representable by it. And "experiencing" merely means that we are processing those representations in a way that enters our cognition, like "mere" perceptions do as well.
It seems that you are thinking that perceptions lead to experience, and then nothing else. My view is that perceptions lead to internal processing that in turn is itself perceived by our cognition, in addition to the original perceptions. And it's the texture of this perceived processing that we call "experiencing". That is, it doesn't stop at the "experience" step, because then we wouldn't know anything about it. Instead, the information about the experience then enters our cognitive processing as a subsequent step. And that's how we don't perceive just "red" full stop, but also an associated "experience" of the red. But again, we wouldn't know about that experience if that wasn't information that enters our mind. And I see no puzzle about such information entering our mind.
It seems that you are thinking that perceptions lead
to experience, and then nothing else.
My view is that perceptions lead to internal processing
that in turn is itself perceived by our cognition, in
addition to the original perceptions
If we decide that a simple light sensor isn't "conscious" because it's a pure input/output machine with no machinery that could reasonably be described as "internal processing", then what about a Roomba?
I also don’t think consciousness is black and white, there is certainly a spectrum. To us, qualia have complex, multi-faceted and multi-layered textures, in line with what we can cognitively perceive and process, and they trigger just-as-complex associations and emotions. It seems to me that this richness is a large part of what makes consciousness wondrous.
I don’t think a Roomba is observing its own internal processing. A program that debugs/profiles its own execution and uses those data as inputs of its main functions, or a neural network that feeds back the changing weights of its edges as additional inputs for itself, would maybe come closer to asking that question.
As an analogy, does a simple light sensor qualify as “seeing”? “Seeing” on the level of a mammal, with object and motion recognition and building an internal spacial model of what it sees, along with predicting what will happen in the next moments based on what it sees, is certainly on a whole other level.
What I’m imagining is that the registering mechanism would itself have sensors placed on its wires that measure the currency levels on those wires, and have the measurements of those sensors as additional inputs into the cognition automaton. And then have sensors on the gates and wires of that automaton, which again feed as additional inputs into that same automaton. Add memory and timing delays. And then multiply all that some million times to get to the level of complexity of our inner mind, of our sensory and movement apparatus, and of our mental models.
When introspecting myself, I don’t see or feel or think anything that couldn’t be explained by such a setup. The different textures (qualia, if you will) of what I perceive in my mind have a certain complexity, but that is merely quantitative and structural, not qualitative.
I therefore simply do not agree that there is a hard problem of consciousness to begin with, in the usually given sense. I don’t agree that there is a qualitative difference between the perception of “qualia” and other perceptions. “Qualia” are just a perception of representations and processes happening in my brain. I see no puzzling mystery that would require solving.