One very persuasive but not very coherent intuition is that of an observer in the brain. People want to think that there's some guy or gal sitting in the head somewhere who receives information; like your soul or whatever.
This of course doesn't make a lot of sense under physical causal closure because then you'd have to implement that observer, and that just gives you exactly the same problems you already have with explaining how action and consciousness work. So you don't gain anything.
Then others want to toss out the idea and with it the idea that there is a center to consciousness. But that's just as bad of an error. The fact that our experience is unified (the boundary problem) is one of the most important hints to how stuff works.
These two ideas need to be uncoupled. The observer is bogus, but the center is not. So if you perceive a unified visual field, there *is* a physical spatial thing in the brain corresponding to the field. But it's not like the field itself does nothing and someone is observing it since again that only pushes the problem down one level. Rather, the field itself is doing computationally relevant stuff, and that is where the consciousness comes from. Said stuff includes all the Gestalt laws: continuation of lines, grouping of objections, 3d construction based on 2d stimuli, and so on.
Ipso facto, the feeling of "observing" a unified visual field is nothing philosophically deep or fundamental; it's just a particular quale that for some reason is computationally helpful. But the feeling of having one in the first place is both of those things.
There can then be also be processes taking the finished product and using it to do stuff -- ultimately to compute motor commands -- but that part has to be unconscious.