1 of 38

�General Intelligence System�

Building a Knowledge-based General Intelligence System

Michael Molin

2 of 38

General Intelligence System - https://activedictionary.com/

Languages:

Afrikaans, Arabic, Armenian, Azerbaijani, Belarusian, Bengali, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, Georgian, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Javanese, Kazakh, Korean, Kurdish, Kyrgyz, Latin, Latvian, Lithuanian, Malay, Marathi, Norwegian, Persian, Polish, Portuguese, Punjabi, Romanian, Romani, Russian, Sanskrit, Serbian, Slovak, Spanish, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Xhosa, Zulu

Features:

- allows you to organize a detailed semantic analysis of the vocabulary of the text to understand the structure of sentences with the ability to translate both words and phrases, and the text as a whole.

- helps you to systematize and store scientific knowledge in your glossaries and dictionaries.

Technology:

Semantic Multilingual Model - a coherent conceptual model of the world transferred into all the existing languages. The model is represented in wordhippo.com on the level of individual meanings and in powerthesaurus.org on the level of word combinations.

Patents:

Semantic Multilingual Model (patent US20210334470 - Neural Network for Interpreting Sentences of a Natural Language) - https://patentscope.wipo.int/search/en/detail.jsf?docId=US339762244&_fid=WO2020106180

3 of 38

What is General Intelligence System?

Universal Principle of Transformer: NLU - NLP - DL - RLHF

NLU (Semantic Multilingual Model (activedictionary.com), Semantics, Knowledge) –

Multilingual NLP (Statistical Language Model (LLM), Syntax, Word Forms) –

Multimodal DL (Morphology, Text, Audio, Images, Video, Interaction) –

Reinforcement Learning from Human Feedback (Pragmatics, Knowledge for NLU)

Translation is an emergence of understanding.

Language is a general human intelligence expressed in symbols and semantic relations between them and a Semantic Multilingual Model is a General Intelligence System.

4 of 38

The First Principles of Nature

  • Molecular Intelligence Results in Chemical Bonds, Genetical or Cell Intelligence Results in Protein Structures and Human Intelligence Results in Human Language

  • Mendeleev's Periodic Table of Elements is a Chemical Alphabet of Nature serving as a World Model, DNA Code is a Genetic Alphabet of Biological Life and both are like a Japanese Hiragana Alphabet consisting of syllables or Chinese Table of Hieroglyphic Keys serving as a Foundational Model of Language

  • A feature of the Periodic Law among other fundamental laws is that it does not have an expression in the form of a mathematical equation. The Periodic Law is universal for the Universe and is the discovery of the interconnection of all atoms in the Universe.

5 of 38

General Concepts

Time is the actualization of space. (in English)

Время - актуализация пространства. (in Russian)

Space - Time - Energy - Interaction →

Structure - Stream - Result - Way

Matter - Projection - Energy - Interaction

Protons/Neutrons - Photons - Electrons - Chemical Bonds

Nucleus - Wave - Quantum - Interaction

DNA - RNA - Protein - Signal cascades

Mass - Light - Energy - Interaction

Knowledge - Consciousness - Understanding - Memory

Intuition - Thinking - Sensing - Feeling

Meaning - Information - Function - Learning

Space - Gravity - Quantum - Interaction

Basis - Process - Result - Way

Dark Matter - Universe - Dark Energy - Interaction

6 of 38

Epistemology - Knowledge is Language

  • Наука, изучающая знание как таковое (от др.-греч. ἐπιστήμη «научное знание, наука», «достоверное знание» + λόγος «слово», «речь»)
  • Intelligence – Знание, Полное Понимание Концептов (holistic approach, живой язык, эволюционное развитие, понятная, последовательная структура описания)
  • General Intelligence System – Знание Природы - (General Intelligence – генофонд природы, гены-концепты – инстинкты - a priori knowledge), Опыт Человека - (Art, Science, Culture, Society – живой язык общения - долговременная память - a posteriori knowledge – эпигеном – регулирование экспрессии генов)

7 of 38

Epistemology - System of Knowledge

Происходит от англ. epistemology «эпистемология», составленного из epistem- + -logy. Первая часть — из др.-греч. ἐπιστήμη «умение; знание; наука», далее из ἐπί «на, при, за, согласно», из праиндоевр. *op- «сзади, вокруг», + ἵστημι «ставить», далее из праиндоевр. *sta- «стоять». Вторая часть — из др.-греч. λογία «сбор пожертвований, подаяний» от др.-греч. λέγω «говорю, сообщаю, рассказываю»; родств. λόγος «слово, речь, разум; мнение»; восходит к праиндоевр. *leg- «собирать».

8 of 38

General Intelligence System Goals

Modeling Human Experience - World - Life - Language

Understanding - Sensing - Protein

Con-sci-ousness - Thinking - RNA

Memory - Feeling - Epigenome

Knowledge - Intuition - Genome

9 of 38

General Intelligence System

Multimodal DL

Восприятие

Понимание

Мир

Multilingual NLP

Мышление

Сознание

Жизнь

NLU

Память

Опыт Человека

Живой язык

Reinforcement Learning from Human Feedback

Взаимодействие

10 of 38

General Intelligence System Principles – Step 1�AI-completeness theorem - Solution�

If you can solve one problem in the AI domain, then you can solve them all.

Marvin Minsky

Language is a structure of knowledge and experience - on a cell level it's a human genome (DNA) and epigenome (histones) - Nature is a great teacher.

Modeling the structure and code of Human Genome-Epigenome (nucleotides of DNA + proteins of histones synthesized by translation) = Modeling a Human Language through the process of translation of two languages

Nucleoproteins (DNA + proteins of histones) - 2 languages - language of nucleotides (5 elements) and language of amino acids (20 elements) = 2 languages of translation matrix

20 thousand human genes encoding proteins = 20 thousand key concepts in the language - a basic vocabulary of a human - 20 thousand words - a basic translation matrix

Epigenetic markers of Genome for epigenetic memory = coding of the translated meanings of the matrix (1 - General meaning, 0 - Options) when assembling a pre-translation model

Double structure of DNA for cell division = Transfer Learning of Knowledge using translation

11 of 38

General Intelligence System Principles – Step 2 – Causal relations

Genetic relations are equal to causal relations

Генетика (от греч. γενητως — «порождающий, происходящий от кого-то») — раздел биологии, занимающийся изучением генов, генетических вариаций и наследственности в организмах, наука о генах.

12 of 38

General Intelligence System Principles – Step 3�Central dogma of molecular biology

Геном

Знания

Сигнальная трансдукция – Взаимодействие

An RNA-based theory of natural universal computation - Hessam Akhlaghpour - https://youtu.be/pcEod3-o_pg

Intelligence => Your Memories of the Past + Your Imagination about the Future = Your Decisions in the Present => Recording Memory => The Untold Story of Sanskrit || A film on research about Sanskrit & Computing || Project SHIVOHAM - https://youtu.be/eQqJkm_q93k

Белки

Восприятие

РНК

Мышление

Эпигеном

Память

RNA Processing = Internal modeling = Transcription - Splicing - Translation - Signaling ->

Intelligence = Internal modeling = Analysis - Search - Synthesis - Research ->

13 of 38

General Intelligence System Principles – Step 4

Principle of relativity for time

Время неотделимо от наблюдателя (время неотделимо от трех пространственных измерений и зависит от скорости наблюдателя)

14 of 38

General Intelligence System Principles – Definition

Principle of relativity for consciousness

Мышление неотделимо от носителя (знания природы-геном-инстинкты - опыт человека-культура-эпигеном - его выражение - мышление)

15 of 38

Rules of consciousness

Закон 2 - 4-20

Системы представления концептов -

предметная, внутренняя, имплицитная память, знание, концепты - 4 буквы нуклеотидов ДНК - интуитивный выбор (кратковременная память оперирования предметами-концептами - 4 предмета-концепта - частота слов с количеством слогов в пределах четырех составляет 90-99% в различных языках);

образная, внешняя, эксплицитная память-отражение - со-знание - 20 букв кодирования белков - количество букв в естественном языке (латиница 26 - 6 парных согласных - 20)

Закон 1 - Закон сохранения смысла

Количество концептов или корней слов в языке равно количеству генов, кодирующих белки - 20 тысяч

16 of 38

Rules of consciousness

Закон 3 - Закон сочетаемости в языке

Количество словоформ и устойчивых словосочетаний в языке равно количеству известных белков - 200 миллионов.

17 of 38

Principle of conservation for consciousness - Definition

“В начале было Слово - Бытие.”

“Логика - принцип сохранения мышления - Гегель”

Слово - последовательность букв языка, выполняющая определенную функцию. 

Ген - последовательность букв генетического кода - нуклеотидов, выполняющая определенную функцию.

Principle of conservation for energy

Фундаментальный закон природы, установленный эмпирически и заключающийся в том, что для изолированной физической системы может быть введена скалярная физическая величина, являющаяся функцией параметров системы и называемая энергией, которая сохраняется с течением времени. Поскольку закон сохранения энергии относится не к конкретным величинам и явлениям, а отражает общую, применимую везде и всегда закономерность, его можно именовать не законом, а принципом сохранения энергии.

Principle of conservation for consciousness

Количество генов, выполняющих определенные функции кодирования процесса синтеза белка, в качестве наглядного примера, участвующего в качестве клеточных рецепторов в процессе восприятия - зрение, слух, осязание, обоняние, вкус, пространственная ориентация - важной составляющей процесса мышления, жизни организма, сохраняется для ее поддержания и равно количеству концептов языка - функций объектов, понятий, представлений об окружающем мире.

18 of 38

Semantics

Shared Culture

Shared Language

Shared Interest Domain

Writer’s

Intent

Reader’s

Understanding

Document

Sentence

Word

19 of 38

NLU - Neural Network for Interpreting Sentences of a Natural Language

US20210334470, CN113366484, RU0002699396, WO/2020/106180

A Semantic Neural Network

20 of 38

A Semantic Neural Network

The invention relates to computer systems and is intended for interpreting sentences in natural language. A neural network consists of layers of neurons. A first neuron layer is intended for interpreting the meanings of individual words in a sentence. Intermediate neuron layers are intended for interpreting word groups in a sentence, wherein the number of words in a word group corresponds to the number of the neuron layer. The last neuron layer is represented by an output neuron and is intended for interpreting a sentence on the basis of the word meanings interpreted in the preceding neuron layers. Each neuron of the input and intermediate layers comprises a memory cell intended for storing the meanings of a word or word group. Moreover, in order to improve the quality of interpretation, each such meaning is assigned a binary attribute, the coding of which is described according to the rule that “1” is the primary meaning of a word or word group and “0” is a secondary meaning of the word or word group.

21 of 38

General Concepts

  • NLU (Semantic Multilingual Model) is about conceptual understanding by human intuition of the words of language and creating world knowledge

  • Multilingual NLP is about statistical processing of this human knowledge and automatic generation of the derivative forms of the words in all languages.

  • Multimodal DL provides an interface for interaction between humans and the General Intelligence System.

22 of 38

Reinforcement Learning from Human Feedback

This approach to DL is also called augmented intelligence or human-in-the-loop and assumes people and technology are working together. This means certain tasks are completed by an algorithm and some by humans. Also, people can take over a process when the General Intelligence System has reached the limits of its capabilities.

23 of 38

A Unified Theory - Universal Language

Definition:

Knowledge - Con-sci-ousness - Understanding - Memory → (in English)

Знание - Со-Знание - О-Сознание - По-Знание (In Russian)

(in Hebrew) כתר - חכמה - ‏בינה - דעת

Unified Theory is Language.

Unified Theory is World Model.

Reality - Illusion - Truth - Way

God - Reflection - Understanding - Way

Nature - Beauty - Truth - Way

A Universal Principle of Transformation:

Life is Evolution.

Atom - Cell - Organism - Planet →

Planet - Solar System - Galaxy - Universe →

Simple - Complicated - Complex - Chaotic →

24 of 38

Transformers for Science:

Meaning - Information - Function - Learning →

Memory - Indices - Heuristics - Rewards (Alan Turing)

Concepts - Categories - Representations - Environment

Knowledge - Reasoning - Inference - Cognition

Analysis - Search - Synthesis - Research

Reality - Theory - Implementation - Practice

Transformers for Physics:

Space - Gravity - Quantum - Interaction

Structure - Stream - Result - Way

Space - Time - Energy - Interaction

Protons/Neutrons - Photons - Electrons - Chemical Bonds

Mass - Light - Energy - Interaction

Base - Emitter - Collector - Signal

Nucleus - Wave - Quantum - Interaction

Structure - Force - Result - Way

Earth - Air - Fire - Water

Matter - Projection - Energy - Interaction

Basis - Process - Result - Way →

25 of 38

A Universal Cycle

Dark Matter - Universe - Dark Energy - Interaction

Universe is Light

Transformers for Technology:

Semantic Multilingual Model - Statistical Multilingual Model - Multimodal DL - Reinforcement Learning from Human Feedback

General Model - Training from scratch - Fine Tuning - Retrieval Augmented

* General model is language itself

Core - Processor - Interface - Human

Data - Encoding - Decoding - Training

Library - Keyword - Text - Context

NLU - NLP - Multimodal DL - RLHF

NLU (Semantic Multilingual Model, Semantics, Knowledge, activedictionary.com) - Multilingual NLP (Statistical Multilingual Model (LLM), Syntax, Information, Word Forms) - Multimodal DL (Morphology, Function, Interface) - Reinforcement Learning from Human Feedback (Pragmatics, Memory, Knowledge for NLU)

26 of 38

Transformers for Biology:

Cell Model

Genome (DNA): Epigenome (Histones (Protein) and DNA) - Transcriptome: pre-mRNA: mRNA: Exons (Splicing): Introns (Alternative Splicing) - tRNA (Translation): Protein - Ca2+ Signal cascades

DNA - RNA - Protein - Signal cascades

Transcription - Splicing - Translation - Signaling

Brain Model (Organism-level Model)

Glial Meshwork - Neural Network (Synapses and Axons) - Neuromediators - Blood Vessels →

Experience - Modeling - Understanding - Memory

27 of 38

Transformers for Linguistics:

Source Terms - Target Terms - Translation - Human →

Meaning - Scheme - Translation - Human -→ (in English)

Смысл - Схема - Перевод - Человек (in Russian)

Root - Model - Word - Context →

Noun - Adjective - Verb - Preposition →

Thesaurus - Ontology - Taxonomy - Relations →

Personal - Artificial - Collective - Multilingual →

Knowledge - Information - Data - Wisdom →

Structure - Stream - Understanding - Memory →

Text - Reading - Understanding - Context →

DNA - RNA - Translation - Signaling →

Semantics - Syntax - Morphology - Pragmatics →

* A word gets its meanings from analysing all the contexts in which the word can appear.

Semantic Multilingual Model (Translation Matrix, Relational Database, Knowledge, Meanings, activedictionary.com) - Semantic Search Engine (LLM, Word Embeddings, Information, Word Forms, google.com) - Translation - Human →

Knowledge - Information - Understanding - Memory →

Translation is an emergence of Understanding.

28 of 38

Transformers for Psychology:

Intuition - Thinking - Sensing - Feeling → (Carl Jung)

Orient - Decide - Act - Observe → (John Boyd)

Subject - Objective - Realization - Society →

Conclude - Plan - Do - Reflect → (David Kolb)

Reality - Illusion - Truth - Way →

God - Reflection - Understanding - Way →

Family - Love - Child - Society

Nature - Beauty - Truth - Way →

“You can't solve a problem on the same level that it was created. You have to rise above it to the next level.” - Albert Einstein

The key to applying the Unified Theory is a system approach - there are seven levels of entities - atom, cell, organism, planet, solar system, galaxy, universe - that use the same structure for transformation or life cycle and entities at one level exist and work in parallel and are replaceable for implementing different functions to make an entity of next level e.g. cells are forming an organism.

Any entity is structured as a transmitter-receiver system and it has four components - a core, input and output channels for information transmission and an interface with the world.

29 of 38

Transformer principle is universal and it resembles an intuitive technology for translation of information into function that forming the universal structure or language that is universal interaction.

Language is any interaction or a process or a life cycle of all entities and that is manifested in a fractal geometry of the Universe e.g. in the similar forms of galaxies and snail's shells.

Nothing is outside language it's reality and it's a universal mechanism or universe - the transformer principle realised in fabric of universe and this principle works at all seven levels of entities. Gravity is its manifestation and maintaining its wholeness.

World Model = Language Model = Universal Laws = Nature is a fractal that repeats in structure - atom, cell, organism, planet, solar system, galaxy, universe. Any entity (not its part) is structured the same way at any level of Nature. All cells in an organism work in parallel.

Foundational Model of Intelligence

Intelligence = Internal modelling = RNA Processing = Transcription - Splicing - Translation = Analysis - Search - Synthesis - System 1 is signal transduction (NLU) and System 2 is a DNA-RNA-Protein way (NLP).

Epistemological General Intelligence System = Knowledge-based General Intelligence System = Neurosymbolic Language Model = NLU - Multilingual NLP - Multimodal AI - Reinforcement Learning from Human Feedback

Relativity Principle for Consciousness

Consciousness cannot be separated from its carrier.

30 of 38

Levels of Entities and Intelligence - Gene Regulatory Network -

OSI Model of Network Architecture - Structure of an Entity

== Membrane of Cell == Protein receptors == Skin == Magnetic Field of Earth == Universe ==

7 - Protein - Universe - Application layer

High-level protocols such as for resource sharing or remote file access, e.g. HTTP.

6 - Translation - tRNA - Galaxy - Presentation layer

Translation of data between a networking service and an application; including character encoding, data compression and encryption/decryption

5 - Ribosome - mRNA - Star System - Session layer

Managing communication sessions, i.e., continuous exchange of information in the form of multiple back-and-forth transmissions between two nodes

31 of 38

5 - 7 - Sensing - Understanding - Electrons - Energy - Result - Function - Morphology - Representations - Synthesis - Act - Word - Collector - Fire - Interface - Decoding - Multimodal DL - Realisation

Data - 5 - 7

== Nucleus of Cell == Brain == Earth ==

4 - Planet - Transport layer - Segments, Datagrams - Spliceosome - Splicing - mRNA

Reliable transmission of data segments between points on a network, including segmentation, acknowledgement and multiplexing

3 - Organism - Network layer - Packets - Transcription - pre-mRNA

Structuring and managing a multi-node network, including addressing, routing and traffic control

3 - 4 - Thinking - Consciousness - Photons - Projection - Time - Stream - Process - Information - Syntax - Categories - Search - Decide - Model - Emitter - AIr - Processor - Encoding - Multilingual NLP - Objective

32 of 38

2 - DNA - Cell - Data link layer - Frames

Transmission of data frames between two nodes connected by a physical layer

2 - Intuition - Knowledge - Structure - Space - Protons/Neutrons - Matter - Basis - Meaning - Semantics - Concepts - Analysis - Orient - Root - Base - Earth - Core - Data - NLU - Subject

== Nucleus of Cell == Brain == Earth ==

1 - Signal cascades - Atom - Physical layer - Bits, Symbols

Transmission and reception of raw bit streams over a physical medium

1 - Feeling - Memory - Chemical Bonds - Interaction - Way - Learning - Pragmatics - Environment - Research - Observe - Context - Signal - Water - Human - Training - Reinforcement Learning from Human Feedback - Society

== Membrane of Cell == Protein receptors == Skin == Magnetic Field of Earth == Universe ==

33 of 38

Symbolic Multilingual Model is a coherent conceptual model of the world transferred into all the languages. The conceptual model of the world is presented in wordhippo.com on the level of individual meanings and in powerthesaurus.org on the level of word combinations.

General Intelligence System - https://activedictionary.com/

Symbolic Multilingual Model:

Afrikaans, Arabic, Armenian, Azerbaijani, Bengali, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, Georgian, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Javanese, Kazakh, Korean, Kurdish, Kyrgyz, Latin, Latvian, Lithuanian, Malay, Marathi, Norwegian, Persian, Polish, Portuguese, Punjabi, Romanian, Russian, Serbian, Slovak, Spanish, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Xhosa, Zulu

�US2021033447 - Neural network for interpreting sentences of a natural language

https://patentscope.wipo.int/search/en/detail.jsf?docId=US339762244&_fid=WO2020106180

NLU (Symbolic Language Model, Ethics, Concepts)

34 of 38

Transformer - Technology of Meaning Extraction

Tрансформер: транскрипция-сплайсинг-трансляция-сигнальная трансдукция

Tрансформер: NLU - NLP - DL - RLHF →

Tрансформер: Meaning - Information - Function - Learning →

Трансформер: память - процессор - интерфейс - устройство ввода

данные - кодирование - декодирование - обучение

транскрипция - сплайсинг - трансляция - сигнальная трансдукция

ДНК - РНК - белки - сигнальные каскады

интуиция - мышление - восприятие - чувство

анализ - поиск - синтез - наблюдение

знание - сознание - понимание - память

35 of 38

The Functions of Consciousness - Carl Gustav Jung

36 of 38

Трансформер для изучения языков

Language Learning:

Text/Imagination - Reading/Writing/Speaking - Grammar/Understanding - Listening/Feeling/Context

текст/воображение - чтение/запись/говорение - грамматика/понимание - восприятие на слух/ощущение/контекст

Linguistic Resources Hub - 64 languages

https://michaelmolin.substack.com/p/linguistic-resources-hub-64-languages

37 of 38

References

38 of 38

General Intelligence System - https://activedictionary.com/

Translation is an emergence of understanding.

Natural Language Understanding - https://www.linkedin.com/newsletters/natural-language-understanding-7020382304715911168/

https://michaelmolin.substack.com