The CPU (or rather, the entire computer) isn't magically generating out of nothing something new. In a certain sense, there is no OS on a computer per se and apart from human interpretation. It is what Searle calls the "observer relative". This is in opposition to things which have objective reality apart from human observation. By analogy, look at text on a page. Per se, these are blobs of ink on a page. They are text only by convention: the person who arranged these blobs was using this convention when arranging these blobs, and you need to know the convention to interpret them as text.

Too often, "emergence" is one of those "gap" explanations that tries to produce something from nothing. Yes, independent things can enter into interactions and relations that result in states of affairs that they alone could not have produced. That's obvious. You don't get a pile of oranges without a bunch of oranges. But a state of affairs of individual objects doesn't transcend being a state of affairs of individual objects as long as you maintain that they are still individual objects.

> they seem to be advocating for some sort of dualism.

This paper is doing the opposite. It's arguing that other papers/ideas advocating dualism are going the wrong way and purely physicalist explanations are the best path forward.

Purely materialist explanations, you mean? Platonic dualism, as advocated by folks like Max Tegmark, isn't non-physicalist. And it needn't be dualist even — if the world is made of math and the materials emerge by implication.

What does it mean to be "made of math"? This reads like gibberish to me. Mathematics deals with formal abstraction. Geometry, for example, abstracts the spacial characteristics of matter from matter and focuses on them solely while ignoring everything else. (This actually reminds me of Bertrand Russell's structuralist account of physics and how it omits much if not most of reality.)

(Also, Platonism does posit an immaterial realm of the forms in which all material things participate. But this participation relation is problematic, something Aristotle pointed out in the Third Man Argument and something Plato himself knew.)

"Gibberish": Plato, Russell, Penrose and Tegmark?

Made of math is as Aristotle noted: "All is number" — Platonic-Pythagoreanism was at the heart of classical civilization and the enlightenment.

I can only suppose that there is something about these ideas that deeply disturbs or even frightens. But I wouldn't dismiss them as gibberish.

The things that happen when a program execute are all emergent behavior, but it's behavior that is well understood and we know everything about how it emerges.

It's all just that, a "behavior". Things that move. Signals propagate across wires. Pixels light up. All the is well understood.

Consciousness is nothing like that.

Why not?

Series of complicated bio-electrical interactions? Vast numbers of connections, neurons firing together?

Even our limited attempts at neural networks are not very well understood, and that is just code running on well understood computer. A brain is way more connected and complicated than something like deep mind, even if you think of that stuff like a first order approximation. So if we cant easily do that, seems an actual brain with way more recurrence and structure is much harder.

I think there are experimental results though, we can look at an brain scan and nearly deduce what people are looking at, although don't quote me I would have to find the paper. Seems pretty physical.

To respond to another part of this, it doesn't matter if we can look at a brain scan and predict perfectly exactly what the scanned subject is thinking. That only answers the question of "how do thoughts occur" not the question "what are thoughts?"

We have no way of even constructing the concept that gets around this. "This brain state corresponds to these thoughts." Okay, but where/what are the thoughts? In order for something to corresponds to the thoughts, they must exist in some capacity, right? So long as consciousness exists at all, which anyone who experiences it can say with certainty that it does. If I drink a beer, I feel a certain way. Neurochemically, we understand exactly why this is happening. What we don't understand is how there are "ways to feel" in the first place.

Understanding how objects interact with one another doesn't answer the question of what those objects are; they just are. Understanding the effects of electromagnetic force doesn't answer the question of what electromagnetism is; it just is. With objects, we can actually break things down into a small number of basic components (particles) that depending on their organization make all objects. But these particles are already a thing with no reason; just an axiom we've been able to use to get a mostly logically consistent view of objects.

Consciousness, we have been totally unable to break down into anything. We can see evidence of it in others, and feel it in ourselves, and we can understand how to make it seem to go away, and also what seems to bring about certain effects in it's space (red, happy, warm, salty, etc.), but our understanding is not of those effects--it's only of how a certain material arrangement seems to bring them about.

Because the web application is only an emergent phenomenon once consciousness has already entered the equation. Without an observer, the website is nothing more than the sum of its parts; only we view it as something else. The CPU, machine code, all of the I/O mechanism, eventually just create an illuminated image on a screen that is only its emergent whole when viewed by a conscious observer who sees it that way.

Far from having an even remotely non-referential understanding of consciousness, we don't even have a non-referential understanding of the referent website as it exists in our perception--it just comes back to the same questions that a lot of physicalists seem to refuse to even acknowledge. I know that when I view this screen I see what I see as red, and I know that the material of my brain and the screen are responsible for that, but that does nothing to address what the referent red is in the first place.

How can consciousness be an emergent construct when emergent constructs are only identifiable as distinct from the sum of their parts by making use of consciousness?

Your reasoning seems to be:

Since complex interactions produce computations.

Therefore complex interactions produce subjective feelings.

Where's the connection?

Do you understand that the concept of information is just .. anything at all? Literally the arrangement of atoms in a rock can be considered information. The flow of water in sewer pipes can also be considered information. Not only that, but information processing.

It's not information that _we_ are interested in, but it is information none the less.

Why would consciousness only arise from the signals traveling across the brain, but not from water flowing in pipes?

> I think there are experimental results though, we can look at an brain scan and nearly deduce what people are looking at

ok this is all well understood. The brain does information processing and we can sort of tell what's happening in it.

That's not the question at all.

Why does the information feel like something? That is the question.

There is no reason to believe that the CPU is conscious though, and if it turned out to be, we can't explain it. In fact, there is nothing to suggest other people are conscious other than that they're probably the same as you. The only one you can be sure is conscious is yourself.