Comment by Kim_Bruning | Hacker News Reader

Comment by Kim_Bruning | Hacker News Reader

I was just pointing out the other day that emergence is not magic. It's just the complementary operation to reduction. [1]

More concrete: If we take a human brain apart, we end up with neurons and glia cells, (and a bunch of supporting cast). These all consist of phospholipids, proteins, and nucleic acids (and some more supporting cast). These then all consist of atoms, which consist of quantum particles. No matter how much we look around, that's all that we find in the brain.

This is a reductive approach. If you do reduction, it's like you're differentiating a formula where you keep losing terms on the right hand side. At every reduction step something similar happens: you are losing information about how your parts are organized.

When you integrate, the terms you differentiated out before will show up again. What sort of magic is this, you ask? There is no magic, those terms were there before. Now that you're doing the opposite (complementary) operation, they're coming back.

So by reducing the brain down, we didn't find any singular component that provides consciousness or thought or any such thing. Having eliminated those, whatever remains must be the truth. But what remains?

The only thing that we've missed is the information that we threw out when we were doing reduction. Apparently consciousness and thought and image recognition and pain and feelings must arise from the organization of the components. This is what is called the "emergent properties".

* You can try the same operation with reducing a game computer down to single transistors. You'll find that this is insufficient to explain how Mario works. You need to look at how the transistors are organized into gates, and how the gates are organized to form memory and instructions, and then how those in turn can be used to execute programs. [2]

* Another similar experiment might be to try to find out what makes a watch tick. You can certainly take it apart, but at the end of the day you are left with a pile of parts that definitely don't tick anymore. You have to put them back together in a very specific way to ultimately recover the emergent property: ticking.

[1] https://news.ycombinator.com/item?id=34619834

[2] Inspired by https://journals.plos.org/ploscompbiol/article?id=10.1371/jo...

Further reading on just how far down the rabbit hole of degenerate neurological systems goes:

https://www.cell.com/trends/cognitive-sciences/fulltext/S136...

Very nice.

It's interesting to ponder how we actually understand how Mario works. We choose a suitable abstraction layer (in this case, on level of programming languages or at least asm instructions), and try to simulate and "integrate" it.

It's interesting to ponder how those quantum particles which make up a brain might be entangled with other particles elsewhere or elsewhen in this or other universes.

Consider that the appearance of Walter Cronkite on an old-school OTA television screen is not an emergent property of a television. You can disassemble it, inspect it, find no miniature Cronkite inside, reassemble it with the same organization, plug it in, and get nothing but static. You don't get Cronkite until the CBS evening news is broadcast, and the television is powered on, working correctly, and tuned to the right frequency.

Entanglement does not really work when the particles interact with the environment. In the brain this would be the other particles in the brain so any non-local quantum effects from particles outside of the brain would be miniscule at best.

Also the brain has computational capacity on the order of 10^18 if you include dendrite computation and only classical physics. That is a mind boggling number. Personally I don't think it's necessary to involve quantum mechanics in the function of the brain (even intracrainial), mainly because it already has all the computation that it needs.

So, it should be able to create an artificial brain with the exact same organization as your brain and it will basically think and act just like you, shouldn’t it?

Yes, assuming that it "works the same way" after being turned on. That is, it has to simulate nerves growing, dying, forming new connections, responding to nutrients, etc... the same way as the original meat brain did.

> The only thing that we've missed is the information that we threw out when we were doing reduction. Apparently consciousness and thought and image recognition and pain and feelings must arise from the organization of the components. This is what is called the "emergent properties".

I think this is a clever thought experiment. For those of us who think there is a genuine hard problem of consciousness, it won't convince us that physicalism has anything close to an explanation, and I'll try to explain why.

When you dismantle a brain, or indeed a body, you lose something else as well -- you lose the capability of any kind of consciousness to (a) report back on its experiences (no mouth or no neurons to fire), and (b) store memories of those experiences so that when it's rebuilt it can report them back. So depending on the story we tell, there are other accounts of consciousness that are not ruled out by this experiment, such as panpsychism -- e.g., perhaps there's consciousness even in the atomic things of the world, in which case organisation isn't the relevant factor.

Moreover, suppose the organisation is important in the way this thought experiment suggests. What is the explanation for why some kinds of organisation produce consciousness, while other kinds of organisation do not? Say, the sand on the beach vs the component parts of my brain? I suspect the response will be related to the amount of information or entropy present in one vs the other, which is a measurable physical property. But then we wonder, *how* does information or order or whatever other physical property we point to produce consciousness? These all look like correlations -- try and find a pattern and say "this physical thing (organisation, information, order, firings of neurons, etc) is the thing which has consciousness". Maybe it does have consciousness, but *why*? Even if you show us that organisations of physical things in the right way is the place where we always and only find conscousness, why do we always and only find it there? These things are clearly not identical to consciousness/phenomenology/mental stuff (organisation is not consciousness), so why do we impute consciousness onto them?

We impute consciousness onto them because we have that first hand knowledge that we have an inner life, that there is a what-it-is-like to be us, so we have to locate that somewhere. It can't be eliminated (though some thing it can!), and so we have to find room for it. And if you are a physicalist, then you know, per your theory, that the answer must be found in the physical world somewhere. Therefore, in your clever thought experiment, you find your way to track down organisation as the relevant physical property. But then the question remains, how and why does organisation come to be/produce/create consciousness?

Reply on news.ycombinator.com