[ExI] Fwd: Is Artificial Life Conscious?

Jason Resch jasonresch at gmail.com
Mon Apr 25 18:03:37 UTC 2022

(Resending message which I had to  trim below 160 KB)

On Sun, Apr 24, 2022 at 5:37 PM Brent Allsop <brent.allsop at gmail.com> wrote:

> Hi Jason,
> Yes, we've got to at include the ExI list, for Stathis' (and other's?)
> sake.  He is in the more popular Functionalist
> <https://canonizer.com/topic/88-Theories-of-Consciousness/8-Functional-Prprty-Dualism> camp,
> which you, like so many, appear to agree with.

First, I would like you to deeply consider for a moment the question "What
is matter?"

Note that physical theories are silent on this question. All physics tells
us about matter is its relations, functions, operations, equations, etc.
that predict our future experiences of this matter, but no physical theory
ever tells us "what matter is".

This leads to another thought: perhaps there is nothing more to matter than
the relations it embodies. Could matter exist as pure relations without any
fundamental relata? If there are two different theories for the relata, but
both follow the same relations, is either theory then "non falsifiable"
since no test, even in principle, could distinguish between the two
theories when all the relations are the same?

If relations are all that exist, or if they are all that can be said to
exist, then functions representing those relations embody and susume all of
physical reality, including all possible physical realities. The most
fundamental physical theories suggest the fundamental role of information
in physics, with the boldest hypothesizing that ultimately there may be no
more to matter than patterns of information.

"As we penetrate into matter, nature does not show us any isolated ‘basic
building blocks’, but rather appears as a complicated web of relations
between the various parts of the whole."
-- Fritjof Capra in “The Tao of Physics” (1975)

"Now I am in the grip of a new vision, that Everything is Information. The
more I have pondered the mystery of the quantum and our strange ability to
comprehend this world in which we live, the more I see possible fundamental
roles for logic and information as the bedrock of physical theory."
-- John Archibald Wheeler in “Geons, Black Holes, and Quantum Foam” (1998)

"The burgeoning field of computer science has shifted our view of the
physical world from that of a collection of interacting material particles
to one of a seething network of information."
-- Paul Davies in “The flexi-laws of physics” (2007)

"Maybe the relationships are all that exist. Maybe the world is made of
math. At first that sounded nuts, but when I thought about it I had to
wonder, what exactly is the other option? That the world is made of
“things”? What the hell is a “thing”? It was one of those concepts that
fold under the slightest interrogation. Look closely at any object and you
find it’s an amalgamation of particles. But look closely at the particles
and you find that they are irreducible representations of the Poincaré
symmetry group―whatever that meant. The point is, particles, at bottom,
look a lot like math."
-- Amanda Gefter in “Trespassing on Einstein’s Lawn” (2014)

Should these ideas prove true, then information theory, computer science,
and mathematics are promoted to the fundamental theories of reality, with
physical theories being only locally or contingently true relations, they
would be those structures/relations capable of supporting the emergence of
conscious observers as seen from the inside.

The implication of "everything is information" is that manipulations and
processing of information (i.e. computation) would then be the most
fundamental basic building block, not only of physical reality but also of
minds and consciousness.

In my view, Functionalism is less a theory of mind, than it is a theory of
fundamental reality.

> The Molecular Material
> <https://canonizer.com/topic/88-Theories-of-Consciousness/36-Molecular-Materialism> camp
> just considers "philosophical zombies" to be absurd.

We are in agreement on the absurdity of philosophical zombies.

But if you accept that philosophical zombies are impossible, there is a
simple proof of functionalism:
1. Given the Church-Turing Thesis, any finitely describable process can be
perfectly replicated by an appropriately programmed Turing Machine
2. The human brain is a finitely describable process (given DNA which
describes both the human body and brain is ~700 MB, and moreover, the
Bekenstein bound of quantum mechanics implies a finite information content
for any physical object of finite energy and volume)
3. Given (1) and (2) there exists a Turing Machine that perfectly emulates
the behavior of any human mind
4. Turing Machines can be built using any material as a physical substrate
(vacuum tubes, transistors, electro mechanical relays, gears and levers,
billiard balls,  water pipes, etc.) the material is unimportant
5. Given the impossibility of philosophical zombies, a Turing machine
perfectly emulating a human mind must also be conscious, as otherwise a
philosophical zombie could be constructed
6. Given (4) and (5) the material substrate of a mind is irrelevant, the
same consciousness must result so long as the abstract
function/algorithm/information processing of the emulated mind is preserved.

>   Especially since they are normally defined in a way that different
> qualia may or may not 'superven' on top of identical physical reality.
> this is just absurd and is not scientifically falsifiable.  I prefer
> falsifiable theoretical science to unfalsifiable philosophy.  Also, as we
> point out in our video: functionalists are no better than dualists, as they
> separate qualia from physical reality.

What is "physical reality"? -- What does it include, what does it not
include? What is the reality of mathematical objects, of mathematical
relations, of mathematical truths?
I think that question is potentially a prerequisite to answering questions
of the nature of consciousness.

There is strong empirical validation of functionalism, in that if you
assume the independent reality of the integers and their true relations,
you can recover many properties of our observed physical reality from that
"first principle". That is, you can derive physical law purely from
assuming some very minimal elements of arithmetic. I write more about this
here: https://alwaysasking.com/why-does-anything-exist/#Confirming_Evidence

>   Despite how many times I've asked Stathis for a way to falsify his
> theory, he has yet to describe how functionalism may be falsified.

It is true that functionalism cannot be falsified. But not being
falsifiable is a property of every true theory.

> So to me, it is no better than dualism.  All he seems to do is qualia,
> like redness and greenness, aren't possible, because they, themselves are
> substrate on which consciousness is composed.
> And of course I've considered "tetrachromats" which have 4 primary colors,
> and shimp that must have a lot more than that.  As I always say, I pity the
> bi (2 primary colors or color blind people) or even worse the achromatic
> (black and white only) people, and can't wait till I (a mere trichromat)
> discover what not only it is like for a tetrachromat, but what it is like
> for all those 16 primary color shrimp.  A brain like that is what I want to
> be uploaded to, and how many more physical colors could be discovered after
> that????  Even if we discover hundreds, with many thousands of shades of
> each, that is still a long way from infinite.

I agree. I think that is the ultimate destiny of humanity (or any
intelligent species), to become explorers of consciousness itself. All
technological development in progress is aimed at giving us better control
over our own sensory and conscious experiences, as I describe here:

> And you still seem to be missing something when you say: "in any
> conscious state one finds oneself [in], one can only ever know... that one
> state." and the way you talk about computational binding like: "less-than
> or greater-than comparison operations, equality tests".  These kinds of
> comparisons are always done between specific things or facts of the
> matter.  That's what computation abou tobjects is.  Your one composite
> qualitative experience of the strawberry includes both redness and greenness.
> [image: 3_robots_tiny.png]
> While it is true, all of these 3 different systems can function the same.
> In that they can all distinguish between and tell you the strawberry is
> red, or not.  But that is missing the point of the factual qualitative
> differences of each of these, and the physical qualities they are using to
> represent these differences, or the fact that their knowledge is
> intentionally abstracted away from any physical qualities that may be
> representing them in a way that requires a dictionary.  You can't get
> substrate independence, without a dictionary for each different
> representation that may or maynot be representing the ones or zeros.
> Representing knowledge like the fist two does not require a dictionary,
> which is far more efficient than the 3rd, which does required an additional
> dictionary.  The same way software runs faster directly on physical
> hardware, vs running on virtual machines (requires a functional mapping
> dictionary to different functioning hardware).

I think your interpretation of functionalism may be of a restricted from,
in that you may only being looking at only the output of the function,
e.g., "Does this function output 'I see red?'"
But that is not how I view functionalism. I think the implementation is
important, and to get the qualia of a human seeing a red strawberry
requires a function that is isomorphic to the same functions employed by
the human retina, visual cortex, and arguably other brain regions. If a
robot's visual functions are implemented in a manner that replicates the
same functions and information processing as goes on in the human brain,
the robot can't help but feel the same. Otherwise would be to invite
zombies, or perhaps even worse: dancing/fading qualia scenarios.

> Which brings me to the 3rd strongest form of effing the ineffable, which
> was portrayed in the movie avatar <https://youtu.be/Uf9SWvs4beE?t=12>
> with Neural ponytails.  These could function like the Corpus collosum which
> can computationally bind knowledge represented in the left hemisphere with
> knowledge represented in the right.  With a neural ponytail like that, you
> would experience all of the experience, not just half.  If the first two
> systems in the above image (one's redness is like your greenness) they
> would directly experience this difference, just the same as if your left
> field of vision was looking through red/green inverted glasses.  It is
> called "4. the strongest form" of effing the ineffable, because what you
> directly apprehend is infallible or cannot be doubted the same way  "I
> think, therefore I am" cannot be doubted.
I do think future technologies such as being able to directly
integrate/stimulate/measure/alter/and link neurons within our brains will
enable a renaissance in our understanding of consciousness, as it will
become a science each person can perform experiments with and get immediate
results in terms of altered experiences. And while brain linking could
potentially give a shared quale of "red" between two minds, once split
again can the two formerly linked minds trust their memories of what red
was like when they were linked? William James pointed out that we can't
even know we were conscious 5 minutes ago. How then can we know the red we
saw 5 minutes ago when linked is the same red we remember now? It might be
possible, but there I think there are compelling arguments that cast doubt
as well. In any event, I look forward to the future where minds can be
linked as in Avatar.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20220425/71ae74a7/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 3_robots_tiny.png
Type: image/png
Size: 26214 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20220425/71ae74a7/attachment.png>

More information about the extropy-chat mailing list