Growing in an Environment should not call for Consciousness per se. The Consciousness bias. H-LI.

by Casian STEFAN, Principal Researcher at Essentia Mundi AI Lab. Contact: ai-AT-essentiamundi.com / ai.essentiamundi.com
Jun.2024.

Please consider citation with link, if you derive work. Or contact me for collaboration. Thank you!

Seems I can't approach, in a traditional way of dualism, or even take a middle position nowadays on the problem of consciousness.
Applying Occam's Razor, if I have to write more that what it is here, it should be a sign that the thinking on the matter is not correctly framed. Writing books and spending a lifetime of the matter, means to me that the frame of reference taken non existent, it is too murky, the Symbols are too much overlapping.

In the quest of defining AGI, firstly, the position in this article would be by taking a rather simpler approach towards defining the (human) consciousness. And then test if the AGI therm would still stand. Ideally would be to also make a comparison between the different approaches to consciousness advanced by philosophers. I do have no knowledge of other approaches in detail, this is an unbiased, only through my "I" bias article on this matter.

The problem of consciousness has a rich history maybe since the humans began being what they are: started living together, growing a network, growing a stack of brain layers, dreaming, growing a language, formalization, first Symbols development. Which, in itself, was a gradual process over millennia in the Environment.

In a way, with this I am discarding the research thorough the words alone position, like in NLP, already stated that it is not enough. I find that the most relevant, even prevalent and closer to the language and language of thought together is the Symbols Framework. The work through the words alone does not lead to understanding through meaning and communication. Words are static, Symbols are open, evolve (as a function of the web of the network, of the colony.) Symbols would also support a test-bed for consciousness amount across systems. Zoosemiotics, phytosemiotics and general systemssemiotics (including non living), non-biologicalsemiotics (non living class.)

At this point, it is hard to say that there is no such thing as the "consciousness." From the superposition states in quantum mechanics to the Wittgenstein's thought experiment of Duck/Rabbit "what do you see?" we have clues that it's not that easily to be discarded. Even as an illusion, it persists, it exists in a form.
The nature of it, as perceived, and the many, too many words and complex frameworks about it, it suggests a complexity that it still has to be nailed down. But, i argue here, that its state, its concept, its Symbol, as one that is overlapping, has black-box and supervenient properties.

Not only we are denied to see what is really out there (supervenience of the senses to the Environment,) but there is also a bias to a self, the Symbols are processed through the "I" filter. A new layer of supervenience: of the self to the self.
One can say this is an Observer's bias. One then is allowed (or demanded actually) to take the position that, in this way, human consciousness can't be defined through itself, for itself. It is something that is open due to overlapping, and maybe equally important, subject to further (evolutionary, in an Environment) change. And from these reasons alone, it is excluded from (strong) mechanistic interpretation and formalization. I also think, that from this bias, it is definable rather in an "externsive" way, not from the self-perspective, fabricated position.

Well isn't this valid for all that goes through the phenomenology? It seems so. Then how to approach the further development for a framework?

1. Supervenience

A system does not know beforehand after what to look for, as to the "new level" of itself. This level is within the system itself as a byproduct of doing things in Environment, not as priors in knowing what to do next. The Environment supervenes the Systems. As a system living below its Symbols, not able to fully grasp its "roundness", having a supervenient relationship to Environment, the notion of consciousness of itself defining, can be discarded, not a sine qua non of the systems within the Environment (Gaia.)

2. Dynamics

There are a few other positions one can take when it comes on how it came to be: a certain (Brownian) motion, or vibration as the "consciousness" bringer across all systems. See a mechanistic view within the Environment. And take also the position of emergence in complex systems.
Or, as the basic driver for that, the population of cells work together to do things in Environment, for getting somehow more successful in its navigable space. To achieve a synergy. We already have these mechanics in our sub-systems and we do that further on a self-isolation, world self-altering, swarm approach, self-domesticated entities through our Symbols path. Not on a path to fully define ourselves, not on a path to statically say: this is it, this is "consciousness", but on a path of further dynamics.

3. The Observer's bias, a bit more

I assume the position of the Observer, but not the "consciousness" of the Observer, as a definition within itself. I assume here, I can't define human consciousness not because it is complex (meaning, an incomprehensible nexus) but because of the: Observer's bias, a particular "form of life", even incompleteness principles, halting problems, aimless human detours, change the system once defined, self-made new worlds, breaking reality when looking, etc.

4. The Observer of the other sub-Systems

From the framework of Symbols, the super-set defined by humans, we can assume it encompasses a richer web, a broader frame of reference within the Environment, within many other ones, of the other systems, of other entities. I acknowledge that they are tightly glued to their Environment too, so that "if a lion could speak, we would not understand him.(L.W.)" And we can think and imagine "how it is like to be a bat" and then we can take a sub-position and acknowledge the instances of the point of views, of the both cases. Symbols would allow us to frame such cases.

It is possible though, for the consciousness Symbol (concept), to be applied in orders, to ascribe it to other entities, and even try define it for them. For sub-systems that have a lesser Symbols network representation and even a different one. We can work with a network of sub-Symbols that gets projected on the others and choose to define some sum of its properties and recognize it as "conscious" and see its "consciousness". We apply a structuralism/formalism to them.

5. Define consciousness for others. Measurable

From this view, (almost?) anything can be defined having some "consciousness" property in a way, in a degree. On an idealist, transcendental way, I may be able to define it for other systems though, because they are sub-systems that are in principle possible to be subsumed with our Symbols, below and within the Observer's web of Symbols.
A quantitative degree, measured by how entangled the network and how many Symbols it possesses. Here this would be my test-bed for consciousness levels across all (living)systems.

6. Apply a method for the consciousness of self

A broader than human Symbol system should be able to see it in us. How to accede to that level?
An unbiased superposition from a view sub specie aeternitatis. "To view the world sub specie aeterni is to view it as a whole - a limited whole. Feeling the world as a limited whole - it is this that is mystical.(L.W.)"
Achievable if we can accede to the a next level of generative platform, and retain the former level. Now possible due to humanity's retain-able Sphere and new self-made Environment (the Detour view).
Also maybe to a new kind of structuralism that is able to transcends. An unbiased one. Thinking of mortal platforms, new generative layer (like Apps on OS,) a combination of physics and mathematics that could expose a post-emergent view.
Or through a kind of general simulation.

7. Challenging some therms in AI. H-LI.

For some of the definitions of AI, AGI, ASI we can extend the rationale to following points:

a. From the view of consciousness from above: AGI seems a vague therm and even a miss. There would not be the case to research and discover a "general" one. In essence it will try mirror the human one. The more appropriate therm:
AH-LI: Artificial Human-Like Intelligence.

b. AI therm, is for the time being still usable, although "artificial" would have to be discarded too, on the basis of there will be a different intelligence. An intelligence, like the human one, which in order to be there should be "aligned," and it will always have the quality of "family resemblance", as cousin intelligence to ours.
When thinking "artificiality" by the means of its Environment (Hardware, OS, Platforms, Virtual Worlds, Real World Agents) and its implementation and formal structures, that its layers, would be hidden more and more, and would be gradually vanishing as artificial, as not relevant "species", rather as a human-like one.
The more appropriate therm:
H-LI: Human-Like Intelligence.

c. Consciousness. A quest to define it, by itself, would not be a desirable path as seen from the ideas described so far. The phenomenology of self, would not be sufficient to grasp it completely. At best, we will get a slight distorted mirror image, our biased view of it. A simulacrum.

d. For the H-LI, the same evolutionary principles applies, pushing a structural scaffolding, will eventually get us near. Overall, it is a gradual process, when suddenly there are no more questions like "is it conscious, is it artificial intelligence?" An absolute cultural, environmental and belonging to the respective Zeitgeist instance. A "form of life".

e. We will still push the models but this will not happen within a generation. It is a gradual, evolutive process (akin to ours.) And we should take into account that the consciousness property itself will change as we get along with those H-LI systems we develop (we further change our world, as the detour continues). On a new level of the maturity, of the new emergent qualities, we can say we will achieve the, on par with ours, H-LI.

f. The world is the totality of Symbols. With them, we would be able to quantize how much "consciousness" another system possesses.

g. The human consciousness will always be there, unless we are becoming a lesser Symbols system. But we can cast it to lesser  systems and propose a sense of consciousness of these systems. For H-LI we can try push further a formalism that would approximate an idealist concept of it.

8. Once more about the Environment. To Symbols of sub-systems

When seen as a "garden," the bio-mass has its tips and downs in form of various systems. The whole is this fluctuations and evolution operations at different levels and different frame of references levels. The human "consciousness" as a definable property, would ignore this whole, the Environment as the up-bringer of those systems. We changed our environment slightly, we are a "form of life" with self induced changes, on an aimless detour. Slightly at battle with it.

The argument starts from the Environment. The levels of operating, the frame of references of various systems bring-ed up, the impossibility that a system is able to fully characterize itself, or systems above it. The symbols, as a means of operation of such systems but not of defining itself.
The observer capable of operating with its level symbols, is only able to define what is beneath its Symbols with its Symbols.
A tree, is making the red cherries because it is a means to realize its spreading of seeds scope. It does not operate beyond. No need to "reason" akin the way mammals do. As observer, as of an object below the symbolic complexity of it, it is able to define qualities of that system also as a "form of life."
Has this tree system "consciousness"? Well, it does in the sense in which the Observer has defined the system: life stages, the growth, the navigation of it in an Environment, doing some certain things. An observer sees the system as a functional whole, but based on his Symbols space.

New words:
Systemssemiotics: including non living.
Non-biologicalsemiotics: non living class.
H-LI: Human-Like Intelligence.
_____________________________________
Exploration by C. Stefan, 15.June.2024 [about]
Last update: 15.June.2024 (versions: *) [versions]
"Essentia Mundi" AI Research Lab. [home]
Copyright © 2024 AI.EssentiaMundi.com, all rights reserved.
_
References:
Wittgenstein, Thomas Nagel, Erik Erikson, Charles Sanders Peirce.
The World as the Totality of Symbols [here].
The Environmental Self-isolation Detour of Humanity [here].