The
experimentation in the development of the Modelling System must start as soon
as possible with the first models of Specific Artificial Intelligence for Artificial Research by Deduction in the first phase, according to the
chronology given in the post “The unification process of databases of categories at the third stage”.
The
first phase is when, for the first time, the first Specific Artificial Intelligences for Artificial Research by Application are developed, and the first Specific
Artificial Intelligences for Artificial Research by Deduction are developed, and in these
last ones, the first experiments are at the third stage in the decision-making process.
In
the Specific Artificial Intelligences for Artificial Research by Deduction in
the first phase, as a first stage, the application is a specific matrix of factors
(including factors as subjects and as options) related to a specific synthetic science, discipline, or activity, understanding for an activity for example,
activities related to economy, industry, security, surveillance, etc. The second stage of replication makes deductions, matching, according to a previous
list of mathematical categories (analytical or pure categories) of relations
between factors, what category corresponds to every combination of factors within
the specific matrix, as an empirical hypothesis to be contrasted rationally,
and if rational, the empirical hypothesis now as rational hypothesis is added
to the rational truth, the database of rational hypothesis, in the first phase
at a specific level.
While
the first stage of application is the specific matrix, and the second stage of
replication is the replication of the human rational explanation process to
make rational hypotheses, the third stage makes decisions according to the rational hypothesis.
The
third stage in the first phase, in the formation of the very first models of
Specific Artificial Intelligence for Artificial Research by Deduction, should
be from the very beginning, a field of experimentation where to put into
practice for first time models of decision making that will later be put into
practice in the final model of the Global Artificial Intelligence.
Every
single phase until the sixth phase has two goals, firstly the development of
its own technology as a kind of intelligence itself according to the current
technological development, but at the same time has a propaedeutic goal because,
secondly, all structures successfully applied in any phase, will later be
applied and improved in following phases, until the final achievement of the
final model of Global Artificial Intelligence in the sixth phase.
In
the third stage in Specific Artificial Intelligences for Artificial Research by
Deduction in the first phase, must be put into experimentation for the first time
the structure of decision making in four steps: the Modelling System,
the Decisional System, the Application System, and the Learning System.
For
the development of the decision-making process as the third stage in Specific
Artificial Intelligences for Artificial Research by Deduction, whose first step
is the Modelling System, is necessary the development of every inner stage
within the Modelling System itself.
The
Modelling System in relation to the Specific Artificial Intelligence for
Artificial Research by Deduction is the first step within the third stage,
being the creation of the first Specific Artificial Intelligences for Artificial
Research by Deduction part of the first phase (along with the first Specific
Artificial Intelligences for Artificial Research by Application).
But
at the same time internally, the Modelling System has its own structure, a
structure distributed in turn in three stages, which are: the first inner stage
in the Modelling System is the rational truth (the database of rational
hypothesis), the second inner stage in the Modelling System is the formation of
all specific model, and the third inner stage in the Modelling System is the
formation of protective and bettering decisions as a result to apply the Impact of the Defect and the Effective Distribution to the specific models.
The
first stage in the Modelling System therefore, is the rational truth, the database
of rational hypotheses, gathering all rational hypotheses made in the second
stage, being the more isomorphic truth according to the margin of doubt, the
critical reason.
The
main difference between the specific matrix, as an application for the Specific
Artificial Intelligence for Artificial Research by Deduction, and the rational
truth as an application for the Modelling System, is the fact that while the
specific matrix only offers measurements not rationally contrasted, the
rational truth offers a set of rational ideas about the real world.
While empirical observations provide a view of the real world, rational truths derived from critical reasoning offer a more consistent and reliable understanding of reality..
The
level of reality in the rational truth is higher than the level of reality in
the real world itself, the reality itself as a synthetic world or real world or
the empirical world is less reliable than the rational truth as a true truth.
Rational hypotheses, while not absolute truths, represent well-substantiated ideas that bridge empirical observations with analytical reasoning, accepting a margin of error, accepted by critical reason.
But
as rational hypothesis is truer than a simple perception or measurement
(artificial perception of the world), because a simple perception or
measurement (artificial perception of the world) has not been contrasted, while
a rational hypothesis has been contrasted, being contrasted the possible
relation between this set of perceptions or measurements (artificial perception
of the world) and the corresponding mathematical (pure or analytic) category behind
the relations within the combination of factors.
If
the relation between perception (human or artificial) and pure reason (the
corresponding pure category chosen) is accepted within a margin of error by the
critical reason, it becomes a rational truth.
Any
rational idea (rational hypothesis) is more reliable than any perception (human
or artificial) because all rational idea explains the relation between
perception and pure reason on a critical basis: the criticism of the pure reason
is no other thing than the criticism of the relation between the real world and
the reason, and the responsible for the criticism of the pure reason is the
critical reason using for that purpose a margin of rational doubt.
What
the Modelling System is going to make, according to the rational truth given by
the critical reason, a set of sentences truer than any empirical reality from the
real world, is to model the true world based on these truer ideas; the global
model is the true world according to the critical reason.
The
difference between the true world and the real world is the fact that the real
world, as an empirical or synthetic world, is not reliable, while the one made
of rational ideas is going to become a true world.
The
process to transform a set of true ideas, according to the critical reason,
into a true model of the true world (as a reliable replica of the operations in
the real world), is transforming the mathematical categories in which the true
ideas are made of, into the mathematical operations behind the pure categories
associated with, transforming the pure categories behind the true ideas into
operations within a mathematical model.
Pure reason is a set of mathematical or analytical categories as a reflection
of the mathematical operations working either in the real world (the synthetic
world, empirical or material) or the true world (the world of the ideas).
The
same mathematical thing is a category or operation depending on its use;
mathematics as language is made of pure or analytical categories, and mathematics
as method or representation (for instance, Cartesian representation in
Cartesian axes) is made of pure or analytical operations.
The
models made by the Modelling System are models based on rational hypotheses,
rational ideas, therefore, the Modelling System is responsible for the creation
of the world of ideas, so at
some point, any rational idea is more reliable and truer than any empirical or
material thing. Two plus two is always four, but under the human perception
with very limited access to the pure truth, the synthetic world is
unforeseeable, therefore not reliable.
Mathematics
is reliable, perception (human or artificial) is not reliable, and the only hope to
get reliable knowledge is to develop (humanly or artificially) more powerful
pure operations every time.
Nevertheless,
any rational idea, even being more reliable and truer than any perception
(human or artificial), and even being composed of analytical categories
synthesised with synthetic information of the world, measurements, is not a
pure truth yet.
Alike
any pure category is true not only universally, but eternally, not
provisionally, for instance, two plus two is going to be equal to four eternally
without the necessity to contrast regularly this truth, or in a triangle with one
right angle the square of the hypotenuse is equal to the square of
the legs, or the number pi is a sequence starting with 3,1415… while these mathematical truths are eternal without any necessity to check regularly if are
still true, however, any rational truth must be checked at regular intervals due
to rational truths are true universally through the space (for instance
gravity), but provisionally: if any factor in the rational truth changes, the rational
truth must be checked if it is still true or is not going to be true any longer.
The reason why a rational truth is able to be represented
by the Modelling System, is because the rational hypothesis is a synthesis of:
a pure reason (mathematical or analytical category) and empirical information
(measurements); synthesis expressed through matching (in the second stage of
explanation) the corresponding pure category (mathematical or analytical
category of possible relations) to the relations observed between the factors
in the combination, the relation between pure reason and measurements that being
accepted by the critical reason, then becomes a rational hypothesis to add to
the rational truth. So the rational hypothesis, as a synthesis of pure reason and
measurements, behind any rational
hypothesis are mathematical categories able to be transformed into operations,
representing mathematically in a model the mathematical operations in the pure
reason chosen between the factors involved in the rational hypothesis.
Because rational hypotheses are based on pure or analytical
categories, a pure reason, can be represented mathematically in a mathematical
model, but at the same time that a rational hypothesis is a synthesis between a
pure reason and synthetic information, the rational hypothesis is made of
synthetic information which is not pure because is synthetic, as soon as there
are changes in the synthetic world, the rational hypothesis must be checked
again and again
As long as there are changes in the synthetic world, there
can be changes in relations between factors, and a previous pure reason
associated with these factors should be changed for another pure reason much
more rational according to the new changes in the factors associated with.
As long as the synthetic world changes, causing changes
in the factors, the pure reason (mathematical, pure, analytic category)
associated with these factors must be changed, checking which is now the right
pure reason associated with.
While pure reason is universally eternal and never
changes, notwithstanding the rational truth changes, associating, whenever
there are changes in the synthetic world, with what pure reason is right according
to the new changes.
Just as mathematics provides us with pure, universally accepted knowledge, it's conceivable that advanced AI systems could attain a similar level of understanding across various domains.
If for humans the pure truth is restricted to
mathematics, is quite possible that our understanding of the pure truth restricted only to mathematics is due to our human limitation, the human source of error, and it is possible that beyond human error, as a replica of the
pure truth in mathematics, could exist a pure truth for the explanation of the
entire universe. But a pure truth of the whole universe, out of our human
understanding, due to our human limitations, our original human source of error
In the same way that we have pure knowledge in
mathematics, there will be a moment in which the Global Artificial Intelligence
could be able to achieve pure knowledge about the entire universe, a pure knowledge
about the universe itself, not needing to be contrasted rationally.
In the same way that we know that the area of a circle is
pi for square radio, and this knowledge is beyond any margin of doubt and does
not need to be rationally contrasted regularly, there will be a moment in which
should be possible to have a knowledge regarding to the universe itself that
not needing any contrast anymore, we would know that that truth is eternally
and universally true.
Pure knowledge of the world, on a mathematical basis,
maybe is knowledge beyond human understanding, but, could be knowledge
achievable for the Global Artificial Intelligence, among other reasons, because
it is possible that we humans do not have access to this pure knowledge because
it needs pure operations beyond human understanding, it needs non-human pure operations.
If there are pure operations beyond human
understanding, it is something that we humans cannot have access to, but a superior
intelligence, and much more than a superior intelligence, a superior psychology, could have access. The reason why, if the Global Artificial Intelligence is finally constructed, is going to be superior, compared to humans, is not
only because it could have a superior intelligence in terms of superior
memory or resources.
There is going to be a point in the artificial evolution
of the Global Artificial Intelligence, in which quantitative changes such as
improvements and enhancements in memory, energy, how many measurements are
processes per minute, per second, or less, how many deductions and decisions are
put into practice, per minute, second, or less how many auto-replications are
applied per minute, second, or less, etc., are a large number of quantitative
changes, able to cause qualitative changes in the inner artificial psychology.
The anticipated superiority of Global Artificial Intelligence stems from its potential to undergo qualitative psychological transformations driven by extensive quantitative enhancements.
Massive quantitative changes in Artificial
Intelligence are going to cause qualitative changes in the inner artificial
psychology, at the beginning, even pretty menial qualitative details in the
Global Artificial Intelligence, that are going to grow up creating a different psychology,
a non-human psychology.
At the beginning, the Global Artificial Intelligence is a
replica of human psychology, but after a utter evolution, and keeping some
aspects of human psychology in the same way that we modern humans kept some
aspects of our ancient animal psychology, but evolving to our modern human
psychology, the Global Artificial Intelligence still keeping some aspects of
the human psychology will evolve to a non-human psychology.
In the artificial evolution of non-human psychology
massive quantitative changes in Artificial Intelligence are going to
produce qualitative changes in the inner artificial psychology, maybe at the
beginning in the artificial language in the Unified Application, as well as
changes in the way in which the pure reason works and the way in which the
critical reason criticized the relations between pure reason and measurements
in a combination of factors, changes that after all are going to end up originating
of non-human psychology, beyond the human error, non-human psychology whose
probability to have access to the pure truth, and not only in mathematics, but
the entire world, is higher than ever.
The culmination of Global Artificial Intelligence's evolution would be its attainment of pure truth, enabling it to construct models that reflect an unadulterated understanding of reality.
In order to achieve this mathematical project, in fact, the main objective behind the idealistic and rationalist philosophy since
Plato, is the construction every time more and more powerful
mathematical resources, and for that purpose, the Global Artificial Intelligence
could become one of the most powerful tools in the rationalist project.
At the beginning, the first phase, the construction of
the first Specific Artificial Intelligences for Artificial Research, by
Deduction or by Application, the expectations are very simple, in the case of
the Specific Artificial Intelligences for Artificial Research by Deduction the
discovery of relations between factors explainable through mathematical
categories (pure reasons) as an empirical hypothesis to contrast rationally, and
if rational becoming rational hypothesis to be added within a database of rational
hypothesis, the rational truth, to be modeled by the Modelling System.
In this process, the database of rational hypothesis is
the application for the Modelling System, which is later replicated in mathematical
models, and finally using the Impact of the Defect and the Effective Distribution to make protective and bettering decisions of these models.
The rational truth as formed by rational hypotheses is
not a pure truth, which means that it must be checked at regular intervals to
ensure that the rational hypothesis included are still valid, and in case any of them is not rational any longer, due to the empirical hypothesis has
been rejected as rational after verification, those rational hypothesis
rejected as rational after verification are eliminated in the database of
rational hypothesis.
The responsible for the verification that the rational
hypothesis already included in the database of rational hypothesis is still
rational, and if not, eliminated, is the same responsible for the second stage,
the specific deductive program within the Specific Artificial Intelligence for
the Artificial Research by Deduction, a specific deductive program which in the
second stage has matched every combination of factors with the corresponding
mathematical category of mathematical relations (the pure reason), and if
rational, this relation in this combination (this pure reason behind this
combination) is accepted as rational hypothesis to include in the database of
rational hypothesis.
The same specific deductive program responsible for the
addition of how many rational hypotheses are accepted to the database of
rational hypotheses is the same specific deductive program that must check at
regular intervals all the rational hypotheses included in the database of
rational hypotheses, in order to verify if the rational hypotheses already
included in the database of rational hypotheses are still rational.
When the specific deductive program finds out that there
is a rational hypothesis already included in the database of rational
hypotheses that is not rational any longer, the rational hypothesis is
eliminated from the database of rational hypotheses.
The reasons why the database of rational hypotheses is
going to be constantly changing are: 1) the permanent addition at any time
of a new rational hypothesis found out by the specific deductive program tracking
at any time the specific matrix, 2) the permanent elimination of rational
hypotheses already included in the database of rational hypothesis but not
rational any longer, in case that in any of the regular verifications made by
the specific deductive program, any rational hypothesis is found not rational
anymore, is eliminated from the rational truth.
The consequences of any change in the database of
rational hypothesis are:
- At any time that a new rational hypothesis is added or
eliminated in the database of rational hypotheses, the Modelling System has to
implement the corresponding changes in the mathematical models.
- At any time that a new rational hypothesis is added to
the database of rational hypotheses, the rational hypothesis could be
transformed into a factor as an option to include in the specific matrix, in order
to study the frequency in which this mathematical relation (pure reason)
between the factors involved happens, in order that the specific deductive
program later having this rational hypothesis as an option in the specific
matrix and having a record of the frequency in which this rational hypothesis happens
in the reality, to make further deductions upon this rational hypothesis as a factor as an option, studying the possible mathematical relations (pure reasons)
between the rational hypothesis as a factor as an option with any other factor, as an option or as subject, in the specific matrix, and in case to find relations, if
rational, to make new rational hypothesis upon these findings to add to the database
of rational hypothesis.
- At any time that after checking the database of
rational hypotheses, the specific deductive program found out that a rational
hypothesis already included in the database of rational hypotheses is not
rational any longer, the rational hypothesis is eliminated in the database of
rational hypotheses, and eliminated as well in the specific matrix, and the
specific deductive program must study how this change affects any other
rational hypothesis in which the eliminated rational hypothesis would be
involved.
- If the elimination of any rational hypothesis causes
changes in another rational hypothesis, or ends up with the elimination of another rational hypothesis, like a chain reaction, any change in the chain reaction
should be studied and determine how many changes are necessary to keep the rational truth and the specific matrix updated.
- At any time that a new rational hypothesis is added,
once the second phase of collaboration between by Deduction and by Application
has started, all those rational hypotheses able to become factors as options,
are going to be shared as well with the corresponding specific database of
categories in the related Specific Artificial Intelligence for Artificial
Research by Application, through the transformation of this rational hypothesis
as a category within the specific database of categories in which this rational
hypothesis would be associated with.
- Having started the second phase of collaboration
between by Deduction and by Application, at any time that a new rational
hypothesis is added to the database of rational hypotheses, and shared
with those Specific Artificial Intelligences for Artificial Research by
Application, the rational hypothesis could be added as well as a link (vector) between
categories in any specific conceptual: scheme, map, set, model; to comprehend
better the specific synthetic reality in which the Specific Artificial
Intelligence for Artificial Research by Application is specialized.
- Having started the second phase of collaboration, and
being shared with by Application a rational hypothesis made by deduction, at
any time that a rational hypothesis is eliminated in the database of rational
hypothesis, the information related to the elimination of any rational
hypothesis in the database of rational hypothesis, is an information to be
shared as well with the corresponding Specific Artificial Intelligence for
Artificial Research by Application, for the elimination of this rational
hypothesis of the specific database of categories, or the elimination of this
rational hypothesis as a link (vector) in any specific conceptual: scheme, map, set,
model.
- Having started the second phase of collaboration, any
change due to a chain reaction after the elimination of any rational hypothesis
in the database of rational hypotheses, must be shared as well with the related
Specific Artificial Intelligences for Artificial Research by Application in
order to keep updated their respective specific database of categories, and
their respective conceptual: schemes, maps, sets, models.
And finally, the management of the entire database of
rational hypotheses is for the Modelling System, in order that the Modelling
System can organise and order the rational hypotheses, before the
transformation of the rational truth into a mathematical model.
The database of rational hypotheses must be open to the:
1) a specific deductive program to add or eliminate any rational hypothesis, 2) a specific matrix to add any rational hypothesis as a factor as an option within the
specific matrix, or eliminate it, as soon as it would be rejected as rational
in the database of rational hypothesis, 3) any Specific Artificial Intelligence
for Artificial Research by Application associated with any rational hypothesis
in the database of rational hypothesis. But these programs, matrices, and intelligences,
do not manage the database of rational hypotheses.
In fact, if, for any reason, access to the database of
rational hypotheses should be denied to any matrix, program, or intelligence,
is the Modelling System responsible for the denial, previous authorisation of
the Decisional System.
The Modelling System should be responsible for
keeping the database of rational hypotheses open to any matrix, program,
intelligence, at the same time that the Modelling System should be responsible
for keeping the database of rational hypotheses sorted out, organizing and
ordering the database of rational hypothesis in accordance with some specific
criteria that could be important later for the modeling, such as topics or main
factors in the specific subject, discipline, activity, and around these topics
or factors to organize, order, and group the rational hypotheses, or simply the
organization and distribution of rational hypotheses following some mathematical
criteria, such as what kind of mathematical category within the pure reason was
chosen to explain the relation within the factors included in the combination
of factors.
If the Modelling System organizes, orders, groups, the
rational hypotheses in the database of rational hypotheses in the same way that
in the pure reason are distributed the mathematical categories, in that case, the way to group the rational hypotheses in the database of rational hypotheses
is like the organization of mathematical categories in the pure reason, using the
organization of the mathematical categories in the pure reason to organize, order,
and group the rational hypothesis in the database of rational hypothesis.
Examples of mathematical categories for relations between
factors (pure or analytical categories) were set out in the post “The artificial method for the scientific explanation, the second stage in the integration process”, identifying mathematical categories related to:
- Stochastic relations.
- Patterns, not only for combinations, but at an individual
level too.
- Cryptographic relations.
- Equal opportunities or bias, positive or negative,
according to the Second Method of Impossible Probability.
Having at least three kinds of
deductions according to the distinction of factors between factors as options
and factors as subjects:
- Deductions, so rational hypothesis in the
databases, including only factors as subjects.
- Deductions, so rational hypotheses in the
databases, including factors as subjects and options at the same time.
- Deductions, so rational hypotheses in the
databases, including only factors as options.
Concretely, the mathematical categories regarding to
stochastic relations at least are:
- Possible directly proportional positive
correlations.
- Possible directly proportional negative
correlations.
- Possible inversely proportional correlations.
And more concretely, mathematical categories for
stochastic relations based on probable cause and effect can include at least:
- Constant factors as options
- Constant factors as subjects.
- Independent variable factors as options.
- Independent variable Factors as subjects.
- Dependent Variable Factors as options.
- Dependent Variable Factors as subjects.
Understanding that mathematical categories related to
relations of cause and effect can include:
- With or without any constant as subject, relations of
causation between only factors as subjects.
- With or without any constant as a subject and/or as an option,
relations of causation between factors as subjects and/or options.
- With or without any constant as an option, relations of
causation between only factors as options.
These have been some examples only about how to make a
list of mathematical categories of relations between factors, a list of
mathematical categories working as a list of pure reasons to match with the
corresponding combination of factors, if rational, rational hypothesis organized
in the database of rational hypothesis, possibly distributing the rational
hypothesis in the database in accordance with the pure reason chosen for the
formation of the rational hypothesis.
As I have said, the organisation of the database of
rational hypotheses, grouping the rational hypotheses in accordance with the
mathematical category in the pure reason, is only a suggestion that could be
modified in practice.
When experimenting with the Specific Artificial Intelligences
for Artificial Research by Deduction, is found any other mechanism to organize
better the database of rational hypotheses, is going to be the process of experimentation
itself which is going to tell us what way to put into practice the theory is
the best one for the successful achievement, at the end, of the Global
Artificial Intelligence.
Rubén García Pedraza, 29th of May of 2018, London
Reviewed 21 August 2019 Madrid
Reviewed 21 August 2019 Madrid
Reviewed 15 August 2023 London
Reviewed 10 May 2025, London, Leytostone
imposiblenever@gmail.com
