Re: [ontolog-forum] RDF finally has its long awaited Generic Client!

Dear Alex,

What you want is for me to explain how to formalize knowledge in the
optimal possible way.

As a mathematician,  fascinated by quantum physics, string theory,
astronomy and cosmogeny/cosmology, and the Buddhist Madhyamaka Middle Way
philosophy, which introduces the much misunderstood concept of sunyata,  I
have pondered over how to overcome the limitations imposed by Godel-Skolem
and Turing on formal systems and (recursively enumerable) algorithms.

Skimming through literally hundreds of articles about cognitive
architectures,  neuroscience,  cognition, computer science,  computational
biology, algebraic geometry,  bootstrapping in physics and the theoretical
physics describing the origin of the universe,  I noticed similarities
started to become visible.

Using sunyata as starting point, we can easily see that the origin of the
universe and an empty void all have quantum fluctuations, and that these
also are at work in nature and our brains.

The article MIP*=RE with its corollary proving the Connes embedding
conjecture to be false was the final peace of the puzzle.

It is not mathematically possible to create a formal theory of everything,
not even an approximation of it.

But what we CAN DO is create "confined domains of discourse" in which we
can introduce formalization and consistency.

Now how can we link all of these up into a patchwork, or network of
formalization that together form a formalized whole of all knowledge?

The brain itself provided the answer, and some free thinking about networks
growing randomly and the clustering that will appear as by a stroke of
magic.

Obviously the most intuitive visual conceptualization is by use of graphs,
but there is a catch. So I looked into quantum graphs,  and ended up with
combining the concepts of quantum graphs and co-line graphs.

True to the Buddhist philosophy the new emerging concept is a mandala graph.

To keep it simple, each vertex constitutes a mathematical object with
equations, a special manifold and additional qualities attached, and
mathematical superimposition.

 This construction does not distinguish between physical,  virtual or
abstract objects.

Consequently we can use graph theory, algebraic geometry, category theory
and constructibility theory to create descriptions of quantum physics,
neuroscience, and the field of all mathematical knowledge described by a
single graph network.

It is THIS NETWORK that can be formalized. The beauty of it being
mathematical is that it can sidestep the sticky concepts of time and
causality, which can be introduced if necessary.

In actuality the current state-of-the-art in neuroscience is the closest to
describing a special case of this mandala graph concept.

Now if we look at network theory and how clusters appear naturally we can
intuitively see that the mathematical underpinnings of this network concept
completely do away with the mathematics that defines generative large
language models, which is ever larger scaling high dimensional graphs and
networks and associated tokens.

Scientist work as collaborators in small worlds. And thus knowledge exists
in clusters, but which are highly connected.

So if I may so bold to say, I suspect the trillions of dollars thrown at
the Stargate AI project to be a waste of money, because the mathematics
behind the generative LLMs is flawed.

I recommend you look at:
Https://ve42.co/networksRefs

My article describing this generalized mandala graph concept and its use is
a work in progress.


Milton Ponson
Rainbow Warriors Core Foundation
CIAMSD Institute-ICT4D Program
+2977459312
PO Box 1154, Oranjestad
Aruba, Dutch Caribbean

On Thu, Oct 2, 2025, 04:50 Alex Shkotin <alex.shkotin@gmail.com> wrote:

> John,
>
> I can try to figure out what Milton means if he answers questions (It's
> always interesting to be Socrates). You've chosen the one closest to yours
> from "a continuous infinity of possible starting points."
>
> I write here from time to time that the most commonly used knowledge,
> presented as theories in various textbooks, articles, and lectures, is
> selected for formalization. You can say: don't formalize the Geometry of
> Hilbert, Euclid, or Tarski. And so on for physical theories, and then say
> the same to every computer scientist and ontologist formalizing in RDF,
> OWL2, CL(🤝), Isabelle, Coq, or Lean.
>
> A strange proposition for our community of practice.
>
> Formalization is not the creation of new knowledge. It is the
> formalization of existing, human-verified knowledge for reliable processing
> by computers.
>
> It should be added that we formalize (some would say, crudely, "cram" them
> into a computer) not only theories, but also their models and methods for
> solving problems about the properties of these models [1]. We spend our
> entire lives constructing theories and their models, and testing them in
> practice by solving various problems: Close your eyes and solve the problem
> of taking a sip from your cup of tea.
>
> LLMs show that knowledge can be concentrated, but who better than you to
> know that it can be concentrated in a much more compact and reliable way,
> without any brute force.
>
> Alex
>
> [1] Specific tasks of Ugraphia on a particular structure (formulations,
> solutions, placement in the framework)
> <https://www.researchgate.net/publication/380576198_Specific_tasks_of_Ugraphia_on_a_particular_structure_formulations_solutions_placement_in_the_framework>
>
> "This document describes a specific framework of specific tasks about a
> particular structure posed and solved GNaA Fig.1.1 within the framework of
> a specific theory, namely Ugraphia, the theory of undirected graphs, with
> little involvement of the theory of binary relations, Binria. The task
> framework stores the formulation and solution of tasks in a structured form
> and is intended for use by everyone in the world (be it the world of a
> research group or Humanity): having set a task on the structure before
> solving it on their own, a person can look into the task framework and see:
> perhaps it has already been solved. The structure and tasks about it are
> described in the first paragraph of the first chapter of [GSiA]."
>
>
> ср, 1 окт. 2025 г. в 21:00, John F Sowa <sowa@bestweb.net>:
>
>> Alex,
>>
>> I totally agree with Milton.
>>
>> MP:  The problem here is the implicit discussion about knowledge,
>>  knowledge representation and formal knowledge representation.  These are
>> three distinct layers and because we still do not have a firm grip on the
>> first, which is inextricably linked to consciousness . . . . .
>>
>> I have been saying something very similar to this point again, and again,
>> and again.
>>
>> I'll repeat once more, starting with Milton's point above. For any kind
>> of knowledge representation, there is a continuous infinity of possible
>> starting points and levels of detail or scope.  Every attempt at
>> formalization must make a choice among an infinitely of options.
>>
>> Therefore, the probability that your choice of what to formalize is
>> correct for anybody else is 1 divided by the total number of options -- in
>> other words, 1 divided by infinity.
>>
>> That value is very, very close to *ZERO*.   Therefore, your project of
>> formalization is *WORTHLESS*.
>>
>> So *DON'T *do it.
>>
>> John
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> ------------------------------
>> *From*: "Alex Shkotin" <alex.shkotin@gmail.com>
>>
>> Hi Milton,
>>
>> What do you think about representation of our theoretical knowledge as
>> axiomatic theories?
>>
>> Alex
>>
>> ср, 1 окт. 2025 г. в 18:10, Milton Ponson <rwiciamsd@gmail.com>:
>>
>> As a mathematician I cannot suppress a chuckle here. The problem here is
>> the implicit discussion about knowledge,  knowledge representation and
>> formal knowledge representation
>> These are three distinct layers and because we still not have a firm grip
>> on the first, which is inextricably linked to consciousness,  knowledge
>> representation remains a difficult task to accomplish, and consequently
>> formal knowledge representation, which we are seeking will remain elusive.
>> Large language models ignore the first layer and assume we can use token
>> based systems to create knowledge representation emulation systems that can
>> capture all formal knowledge representation systems.
>> If one looks at the groundbreaking paper MIP*=RE,
>> https://arxiv.org/abs/2001.04383, and what it states about the Connes
>> embedding conjecture being false, this should ring a bell.
>> Because we cannot in all cases assume that a finite matrix in a very high
>> dimensional space can approximate a simulation of an infinite dimensional
>> space.
>> Which means that no matter how high we make the dimension and
>> consequently the number of parameters used, in some cases the simulations
>> will never even get close to approximate a finite accurate model of
>> infinite space.
>> Which means generative LLMs are are a mathematical dead end, and will be
>> the reason why the AI bubble riding on generative LLMs will burst.
>>
>> Milton Ponson
>> Rainbow Warriors Core Foundation
>> CIAMSD Institute-ICT4D Program
>> +2977459312
>> PO Box 1154, Oranjestad
>> Aruba, Dutch Caribbean
>>
>>
>> --
>> All contributions to this forum are covered by an open-source license.
>> For information about the wiki, the license, and how to subscribe or
>> unsubscribe to the forum, see http://ontologforum.org/info
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "ontolog-forum" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to ontolog-forum+unsubscribe@googlegroups.com.
>> To view this discussion visit
>> https://groups.google.com/d/msgid/ontolog-forum/14bb3912000a477f820480fbdd414cbd%40af128af903a246abbaa42dc2aef387a1
>> <https://groups.google.com/d/msgid/ontolog-forum/14bb3912000a477f820480fbdd414cbd%40af128af903a246abbaa42dc2aef387a1?utm_medium=email&utm_source=footer>
>> .
>>
>