Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We Are Beast Machines

nautil.us

Post reply on HN
Print view View original

61–70 of 145 posts

Re: We Are Beast Machines

#61

Post by chmsky00 »

nickelpro wrote: post #3

Nothing in this essay addresses the single most prominent question of contemporary philosophy of mind, which is the hard problem of consciousness. It only addresses the soft problems, which everyone agrees are solvable "merely" with enough time, funding, and initiative. Nothing about the "beast machine" addresses how qualitative experience arises from physical phenomena, or as Chalmers puts it, "Why are we not philos…

The hard problem will be solved in time.

My reasoning is that the people that defined it did so in a time when our science was less developed, so of course the problem seemed much harder.

Information theory makes even the hard problem a matter of understanding interaction fields of physics, with the chemistry of biology. Relativity and sensory network effects explain relative experience elegantly enough.

A lot of theories from back in the day are built on outdated understanding. Unfortunately their authors did not get to see our achievements in engineering unravel a lot of their over baked theories due to a need to fill in gaps without hard evidence. Same as we won’t see technology in the future have no need of all this software we wrote. It won’t literally be handed on.

Luminiferous Aether was once a thing to many even though it’s not one discrete thing. One might consider it was a poetic initial take on field theory, which we now rely on sets of glyphs of shares meaning. Which artistically could also be imagined as flowing sets of matrix code that glow like a luminous field.

If there’s a hard problem to consciousness it’s an unwillingness to consider there is no Valhalla. Is there a hard problem? Or do we hope for an answer that suggests we’re not just meatbags?

Re: We Are Beast Machines

#62

Post by ivanbakel »

nickelpro wrote: post #13
ivanbakel wrote: post #9

Earlier quoted context omitted.

The trouble with the "hard problem" of consciousness is that it's predicated on the possibility of the existence of p-zombies. Even if you don't know how a complex physical system gives rise to consciousness, you understand that a block of steel is not a complex physical system - so the fact that it isn't conscious isn't that contentious. If p-zombies could exist as complex systems without consciousness, then there w…

Really? A sufficiently advanced computer is a perfect P-zombie to me. This is the Chinese Room problem restated. A computer is capable of behaving as a conscious actor without any concept of knowledge or qualia as we would recognize it. But again this gets into a faith-based discussion. To me a sufficiently complex computer doesn't suddenly cross a "consciousness threshold" of complexity, but if to you it does there'…

>This is the Chinese Room problem restated. A computer is capable of behaving as a conscious actor without any concept of knowledge or qualia as we would recognize it.

This is begging the question of p-zombie existence. Searle and Chalmers both think it's possible for arbitrarily complex systems to exist without qualia. But given that qualia is unmeasurable, there's no way to know for sure.

That's why I consider the statement of the problem to be faith-based. It is possible that consciousness is an emergent property, and qualia is experienced by every sufficiently-complex system. Declaring that p-zombies are conceivable requires assuming that this is not so.

Re: We Are Beast Machines

#63

Post by jerf »

dougmwne wrote: post #59
nickelpro wrote: post #3

Nothing in this essay addresses the single most prominent question of contemporary philosophy of mind, which is the hard problem of consciousness. It only addresses the soft problems, which everyone agrees are solvable "merely" with enough time, funding, and initiative. Nothing about the "beast machine" addresses how qualitative experience arises from physical phenomena, or as Chalmers puts it, "Why are we not philos…

To turn this on it's head, maybe the problem is assuming that everything else in the universe is without consciousness aside from certain classes of matter organized in a particular configuration of a brain. If we instead assume that consciousness is an intrinsic property of all matter and space, then the problem goes away. Or to state it another way, either everything is conscious or nothing is.

It doesn't do a darned thing to the problem. It just moves it around; the question becomes "If the entire universe is 'conscious', for whatever your definition is, why are some configurations of matter more able to express it and others less?" How can we build an identifier that says what will be more expressive and less by inputting the configuration? How can we engineer configurations of matter that are more expressive rather than less? It is clearly obviously that configurations of matter are not just additive, that is, every 3 pounds of matter is the exact same type of conscious as any other 3 pounds of matter, so what's the difference, exactly?

Which happens to be the exact same problem.

This problem doesn't go away with any amount of changing definitions. Shifting the mystery under the rug where you can't see it doesn't mean the mystery has gone away, nor that everyone else is going to be fooled. Some of still see the lump under the rug you just created.

Re: We Are Beast Machines

#64

Post by nickelpro »

chmsky00 wrote: post #61
nickelpro wrote: post #3

Nothing in this essay addresses the single most prominent question of contemporary philosophy of mind, which is the hard problem of consciousness. It only addresses the soft problems, which everyone agrees are solvable "merely" with enough time, funding, and initiative. Nothing about the "beast machine" addresses how qualitative experience arises from physical phenomena, or as Chalmers puts it, "Why are we not philos…

The hard problem will be solved in time. My reasoning is that the people that defined it did so in a time when our science was less developed, so of course the problem seemed much harder. Information theory makes even the hard problem a matter of understanding interaction fields of physics, with the chemistry of biology. Relativity and sensory network effects explain relative experience elegantly enough. A lot of the…

The hard problem has technological progress built into its definition. It defines problems of consciousness that can be solved by progress as the "soft" problems, and the problem that cannot be solved by progress as the "hard" problem. The hard problem doesn't say that a mechanism _hasn't_ been found, it says a mechanism _cannot_ be found. Like Gödel's incompleteness theorems or Heisenberg's uncertainty principle it places a limit on what can be known about the system, "[the hard problem will] persist even when the performance of all the relevant functions is explained."

The validity of that is up to you, but if you accept the hard problem as a valid question it will not be solved by technological progress.

Re: We Are Beast Machines

#65

Post by nickelpro »

ivanbakel wrote: post #62
nickelpro wrote: post #13

Earlier quoted context omitted.

Really? A sufficiently advanced computer is a perfect P-zombie to me. This is the Chinese Room problem restated. A computer is capable of behaving as a conscious actor without any concept of knowledge or qualia as we would recognize it. But again this gets into a faith-based discussion. To me a sufficiently complex computer doesn't suddenly cross a "consciousness threshold" of complexity, but if to you it does there'…

>This is the Chinese Room problem restated. A computer is capable of behaving as a conscious actor without any concept of knowledge or qualia as we would recognize it. This is begging the question of p-zombie existence. Searle and Chalmers both think it's possible for arbitrarily complex systems to exist without qualia. But given that qualia is unmeasurable, there's no way to know for sure. That's why I consider the…

I concur with this completely. Rejecting p-zombies as a concept aligns you either with (3) or (4) of the toy framing above, depending on if you're materialist (consciousness emerges from physical phenomenon/doesn't exist) or non-materialist (consciousness emerges from non-physical phenomenon, and does so universally) about the follow on question.

Re: We Are Beast Machines

#66

Post by rmah »

nickelpro wrote: post #6
joeberon wrote: post #4

Earlier quoted context omitted.

So far the only answer I've seen people give to the question is "what problem? There's no problem". Basically, there are many people who don't (or also have convinced themselves that they don't) perceive an incongruity between their personal subjective experience and the idea that consciousness is just an emergent phenomena of a physical system. I haven't found a good way to explain it, I just know that there's somet…

I don't understand how someone can not understand the problem. You and I have conscious experience, we have a meaningful sensation of qualia. A billet of 4140 steel does not, or so we assume. Both are governed by physical laws of the universe, so there must be some distinction beyond these physical laws that differentiate them. In this framing, there are typically four camps, three of them line up with your camps: 1)…

Consider the weather. Weather exists, I think everyone can accept this. But what is weather? Is it simply an aggregate emergent property of many interacting atmospheric and geological systems? Or is it something more? How can emergent properties of even a complex system create self-organizing and self-perpetuating semi-stable phenomenon like tornadoes? What controls the weather?

Consider consciousness. Everyone can accept that it exists. But what is consciousness? Is it really just an emergent property of brain-body activity? Or is it something more? How can an emergent property of a complex system create semi-stable behaviors and self-awareness? What controls consciousness?

IMO, consciousness is just an emergent property of brain-body activity (mostly brain). More recent studies of animal behavior have shown that consciousness is not a binary -- you have it or you don't -- thing. It is a gradient. Different animals possess different levels of cognitive ability, consciousness, self-awareness, social-awareness, etc. In short, there is (again IMO) nothing special about human consciousness: I submit that any sufficiently complex trainable associative-memory neural network (like a biological brain) will possess some level of consciousness as an emergent property. Just like any sufficiently large atmospheric + geological environment will possess weather.

Thus, there is no problem to understand (or not understand).

As an aside, are humans even "fully self-aware"? How would we know if we're not? What does that even really mean?

Re: We Are Beast Machines

#67

Post by unyttigfjelltol »

CuriouslyC wrote: post #49
adriand wrote: post #26

Earlier quoted context omitted.

One obvious difference between computers and humans is that only one of them is an animal. So far as we can tell, consciousness only arises in animals. This sounds like a truism, but I wonder if using phrases like “complex system“ masks it: both humans and computers are complex systems, and only one of them is conscious, and therefore what could possibly be the difference? Well, the list of differences between the tw…

You think your computer isn't conscious? How do you know? Just because it's not demonstrating free will? How many layers of error correction are built into computers to stifle the "randomness" of electronic equipment? Do you think you could still exercise free will locked in a padded room and strapped to a bed with a straightjacket?

This comment is amazing. Thank you.

Re: We Are Beast Machines

#68

Post by nickelpro »

NoGravitas wrote: post #58
nickelpro wrote: post #6

Earlier quoted context omitted.

I don't understand how someone can not understand the problem. You and I have conscious experience, we have a meaningful sensation of qualia. A billet of 4140 steel does not, or so we assume. Both are governed by physical laws of the universe, so there must be some distinction beyond these physical laws that differentiate them. In this framing, there are typically four camps, three of them line up with your camps: 1)…

I agree about the best parties. But what about the idea that consciousness exists on a continuum — that everything that processes information has a scope of consciousness that corresponds to the complexity of what it processes? A block of steel doesn't really process any information, but a thermostat does, and so does a paramecium. I'm happy to grant qualia to a paramecium, even if I don't think those qualia are very…

Asking "how is consciousness quantized?" definitely would put you in (4) and is favored by Chalmers I believe. The idea that consciousness can somehow be "aggregated" by sufficiently complex or anti-entropic systems lends itself to this view.

Re: We Are Beast Machines

#69

Post by Robotbeat »

ganzuul wrote: post #27
pitspotter2 wrote: post #23

Earlier quoted context omitted.

Check out Chapter 11 of The Hidden Spring by Mark Solms. In it he identifies subjective experience, for example perceiving the colour red, as an observational perspective. He reminds us that we can also experience vision objectively, with the right lab equipment, e.g. by listening to spike trains travelling from the retina down the optic nerve. Say we do the latter. We can then legitimately ask what the same process…

Generally it is understood that qualia are not quantifiable.

Isn’t that begging the question?

Re: We Are Beast Machines

#70

Post by slfnflctd »

joeberon wrote: post #32
n4r9 wrote: post #30

Earlier quoted context omitted.

If we take a human and replace one of their neurons with a silicon chip that exactly replicates its electrical activity, is the human any less "conscious"? Presumably their conscious experience is the same (if not why not?) If the human is still conscious, what if we continue one by one to replace all the neurons with silicon chips in the same way? Where does this thought experiment fail?

Personally I think most people who argue consciousness is an emergent property of complicated physical systems would believe that the machine would still be conscious after replacing every neuron with a chip

Given a sufficiently advanced chip which included all chemical interactions (or effective simulations), why wouldn't they be?

If after the transition is complete we determine/decide they aren't conscious, then we have to argue about at what point the 'hybrid' brain ceases to manifest consciousness, and why. Maybe the organic parts would start to change their prior functionality in response to the synthesized parts... but that would suggest we just didn't have a complete enough model of the neuron and we're back to square one. Once that model is complete, there should by definition be no problem, unless you want to argue it can never be complete, for which you then need evidence to convince everyone.

I say it's worth continued study.

Post reply on HN