The World Hypergraph Network: Reality > Science > Technology > AI > Trans-AI = Real AI
“All things are implicated with one another, and the bond is holy; and there is hardly anything unconnected with any other thing.” Quantum entanglement from Marcus Aurelius, Meditations
We introduce three interrelated constructs:
The World Hypergraph Network as Global Causal Hypergraphs of Reality, generalizing Conceptual Graphs, Knowledge Graphs, the Web as the Giant Global Graph for the Semantic Web 3.0, and Graph Neural Networks for AI, ML, DL, ANNs, or Large Language Models.
The World Hypergraph Network embraces the universe of systems, physical, biological, social , information, digital or virtual systems.
Real AI as combining the Reality Modeling Machine (the World Hypergraph Network) + Causality Engines + Data Ontology Engine + AI Models + ML/Deep Neural Networks + Human Intelligence
Trans-AI as a Transdisciplinary, Transformative and Translational Technology designed, developed, deployed and distributed as structured below:
AI = Real AI = Transdisciplinary AI = Interactive AI =
Reality Machine (World Modeling and Reality Simulation Platform, the World Hypergraph Networks + Scientific World Knowledge + the Internet/Web Data) +
Causality Engines (Physical CE, Chemical CE, Biological CE, Mental CE, Social CE, Economic CE, Political CE, Informational CE, Digital CE, Virtual CE, Technological CE) +
PATS [Predictive Analytics Statistic Techniques, Statistical Models, Narrow AI, ML, ANNs, DL] +
[LLMs, Generative AI] +
Knowledge Graphs + Domain Ontologies + Data Sets +
Causal AI (Explainable AI (XAI), "Understandable AI")
Robotic Hyper-Automation +
the Internet of Things +
Emerging Technologies +
Human Intelligence +
Hyperintelligent Hyper-Automation ...
It is driven by the AI Reality Engine, the World Modeling and Reality Simulating Platform, representing the universe as global causal hypergraphs with complex hyperedges, interrelationships, interactions and interdependencies, among and between all hypernodes, entities, states, changes, or causal variables.
Why do We Need the World Model AI?
There are three simple fundamental reasons:
To deeply understand the world for machine intelligence and learning and human intelligence replacing statistical independence with causal world model
To create a deep understandable AI instead of the Explainable AI (XAI)
To build the Trans/Meta-disciplinary AI (Trans/Meta-AI) following the structural algorithm: Transdisciplinary AI (Trans-AI) = the World Hypergraph Network + Data Ontology + AI Models + ML/Deep Neural Networks + Human Intelligence
First, "It takes a human about 20–30 hours to learn how to drive a car, while it takes tens of thousands of hours to train a neural network to achieve this same capability. Even after all of these years of training and despite of using the latest and greatest in processing and sensor technology, self-driving cars are still not deemed road-safe. We can train our learning model to recognize many of these situations, but there is an infinite number of them and even after millions of miles driven, the machine learning model will not have experienced anywhere near all of them.
Why? Because deep learning models do not have an inherent understanding of how the world works. They do not know any laws of physics and neither do they know ethics or even liability laws. Everything they learn is based on statistical independence of all input variables. "
Statistical independence is a key concept in probability theory as well as in ML and DL. Two events A and B are statistical independent if and only if their joint probability can be factorized into their marginal probabilities, i.e., P(A ∩ B) = P(A)P(B). This makes the base for IID random variables, widely adopted in probability theory and statistics, as a collection of independent and identically distributed random variables, where each random variable has the same probability distribution as the others and all are mutually independent.
We go much deeper replacing the statistical independence in the input data with the causal dependence in the input data, as it is prescribed by the Bayes' theorem, Bayes' law or Bayes' rule or Bayes–Price theorem:
Second, ML/AI now competes and outperforms people on many narrow tasks thanks to the development of new learning algorithms, as deep neural networks. These algorithms can discover patterns in large datasets and interpolate their answers when new data is provided without any human interventions. However, they are based on a set of statistical algorithms and non-linear mathematical functions demanding technical explanations to be transparent and reliable.
Explainable AI (XAI) has long been considered a potential solution to publics distrust of AI. There are a lot of scientific studies in XAI has been published every year since the mid-2010s, and there are still critical gaps in the XAI fields. Learning algorithms are trained over various learning algorithms and data types, ranging from numbers to texts and including images and videos. There is no global explanation to make the model transparent and understandable as a whole, see the Classification of XAI methods into a hierarchical system
So, we should shift our focus to "Understandable AI," which can better meet non-technical stakeholders' needs. XAI only approximates the Black Box ML.
Third, to build the Understandable AI is possible if to follow the structural algorithm: Transdisciplinary AI (Trans-AI) = the World Hypergraph + Data Ontology + AI Models + ML/Deep Neural Networks + Human Intelligence
The World Model Network as a Global Causal Hypergraph is embracing all the possible representations and implementations, as Graphs, DAGs, Multigraphs, and Hypergraphs, as well as Conceptual Graphs, Knowledge Graphs, the Web and Giant Global Graph for the Web 3.0.
The World Hypergraphs can be used to model any real-world or virtual reality systems, physical, biological, social , information and digital systems.
The WHN of Changes and Variables
The world is a giant multi-level globally distributed network of variables, as changes, events, actions, and processes.
Its is formalized as a multi-directed global hypergraph of hypervertices/nodes/points and hyperlinks/edges/lines/arrows/arcs/relationships, with its hypernodes as different as in:
Physical variables or quantities
Chemical variables
Biological variables
Human variables (age, gender, health status, mood, background, etc.)
Mental variables
Social variables
Political variables
Economic variables
Cultural variables
Environmental variables….
Variable are a formal concept for qualities and quantities, features or attributes, having categorical and nominal, interval and ration and cardinal levels or scales of measurement.
They number of variables is virtually numberless, all being studied in both functional roles, like vector-valued functions or y = f(x) or x = g (y), or both causal directions, direct and reverse.
Any variable could be studied in any role, as independent variable or dependent variable:
Independent variable causes an effect on the dependent variable.
The dependent variable is dependent on the independent variables, which are controlled, manipulated or altered. And the dependent variable is the variable that is measured or tested by a researcher.
Ideally, one must rely on a causal hypergraph of hyperedges (causal hyperlinks) and vertices (causal variables) of arbitrary size and order in which an edge could join any number of vertices/nodes/variables, where a DAG is a special case.
In an undirected simple graph of order n, the maximum degree of each vertex is n − 1 and the maximum size of the graph is n(n − 1)/2, what is responsible for the network effects.
Science and Engineering, Global Risks, Social Media, molecules, physical interactions, all can be represented as causal hypergraphs (knowledge fields, risk events, users or atoms or particles are hypernodes and interdisciplinary interlinks, causal influences, social relations, connections, bonds or forces are hyperedges).
Below there are three notable examples: Global Knowledge Graph Network, Global Knowledge Graph Network , and AI Graph Network:
Global Knowledge Graph Network
Global Knowledge Graph Network
AI Graph Network as the Patent Landscape
Meta-AI = Trans-AI = Real/Causal AI: Causal World Hypergraph: Artificial Neural Network + Graph ML + Semantic Web + Conceptual Graph + Bayes Network +Global Giant Graph...
Below we show that the best digital application of Causal World Hypergraph (CWH) could be the Meta-AI Web emerging as the next Internet/Web3, dubbed as the Metaverse.
The CWG could be the robust conceptual base and data foundation for the Meta Platforms' AI Research SuperCluster (RSC) introduced as the world’s fastest AI supercomputers to accelerate AI research and help us build for the metaverse.
The CWG is a generalization of graph network models, including
Artificial Neural Networks
Conceptual Graphs
Bayes Networks
Semantic Web
Global Giant Graph
Knowledge Graphs (Microsoft Academic KG and Google KG)
the Graph ML and Graph NN.
We shortly focus on the Graph ML and Graph NN.
Social networks or molecules or physical interactions can be represented as graphs (users or atoms or particles are nodes and connections, bonds or forces are edges).
It is like as the FB social network, with 3 hypernodes and many hyperedges, visualized by Wolfram.
Graph networks or network graphs are a set of data points/nodes and edges/connections/weights; data structures that model relationships between data.
Graph ML is a branch of machine learning that deals with graph data, with nodes having feature vectors associated with them, and edges, which may or may not have feature vectors attached. A feature vector could be 3 XYZ coordinates + some energy and time of arrival information, in case of neutrino detection.
It is represented by Graph Transformers (GTs).
Graph Neural Networks (GNNs) are a class of deep learning methods designed to perform inference on data described by graphs. GNNs can be directly applied to graphs, and provide an easy way to do node-level, edge-level, and graph-level prediction tasks.
GNN directly operates on the Graph structure. A typical application of GNN is node classification, like in Twitter’s GNN to detect/ filter out fake news node by the propagation pattern.
While GNNs operate on normally sparse graphs, GTs operate on the fully-connected graph where each node is connected to every other node in a graph.
What both approaches are missing: any real-world system with defined structure should be modeled by a fully connected causal hypergraph network, not a statistical GML or probabilistic GNN.
The Global Hypergraph and Content Formats Categories
The global hypergraph is underlying content format categories involving sensory experience, model, language or science used for encoding information. There has been a countless number of content formats converting a specific type of data into displayable or observable information:
Domain language formats
Scientific modeling and visualization programming
Natural languages formats
Code formats
Document file format
Audio data encoding
Stereophonic sound formats
Digital audio data
Visual data encoding
Film speed formats
Pixel coordinates data
Color space data
Texture mapping formats
3D display formats
Holographic formats
Display resolution formatting
Text Formats/ Font Size/ Line spacing
Motion graphics encoding
Frame rate data
Video data
Computer animation formats
Instruction encoding
Causal graph data applications are around us: particle physics, computational pharmacology/chemistry/biology, traffic prediction, computer vision, fake news detection, social networks, commercial recommender systems, etc.
Real AI or Causal AI of Machine Intelligence and Learning as the World's Hypergraph Technology
[Real] AI is a great thing of great use, the best strategic general purpose technology with a general intelligence power. As Causal AI, it is included in the Gartner Hype Cycle of Emerging Technology 2022
Regardless all the big tech overhyping, there is no computer intelligence, machine intelligence, artificial intelligence, computational intelligence or real intelligence in existence. It has a honorable status of the greatest human dream, highest ever goal, still residing in the imaginary world of researchers, developers, engineers, artists and sci-fi promoters.
To be discovered, computer’s intelligence, as the greatest ever invention, should pass all the emerging technology stages, as to be: conceived, modelled, designed, developed, deployed and widely distributed.
One might believe that AI is already around us, it is embedded in every parts of our life, in every aspect of modern society.
It assesses, assists, manages, recommends, recognizes, decides, predicts, etc. Playing video games, Google translating, Seeing ads online, Face ID on a phone, Mapping out your destination, Self-driving your cars, Trading your stocks, Playing strategic games, Composing music, Painting arts, Creating stories, and so on and on.
All is good and well, but this all is just automated software with sophisticated statistical and mathematical algorithms, showing no sense or meaning, understanding or intelligence by any good definition, if only small pieces of quasi-intelligence.
"Modern AI systems have made it easy to tackle many problems previously thought out of reach of computers. You have possibly heard of some of these successes such as:
GPT-3: Generates paragraphs of human-like text based upon any initial prompt you provide it.
AlphaFold: Predicts how proteins take shape in 3D space. A true breakthrough in modern biology.
DALLE-2: Creates incredibly detailed and realistic images from text descriptions.
These systems are so good that they have convinced even those working on developing them that they are sentient.
However, despite the successes, many of these systems can be thought of as technological parrots. Parrots can mimic their owners, but do not have a true awareness of what they are saying, nor why they are saying it.
Similarly, modern AI systems can mimic the patterns they have learnt from previous data, without having the true context of the problem which is being solved, nor understanding why a given prediction is returned. Modern AI systems are parrots at both massive scale, GPT-3 was trained on approximately 3 billion web pages, and with huge societal implications.
The end result of this parroting is that modern AI systems suffer from the following three B’s:
Blind
Biased
Brittle
These three B’s mean that modern AI systems are flawed at tackling the nuanced, complex and high risk applications which they are being applied to." [A 6-Minute Introduction to Causal AI]
Being brain-dead, blind, biased and brittle, or "narrow and weak", modern AI/ML/DL systems fail to perform complex tasks of Real/Causal AI, as much as ANNs fail to solve complex problems performed by Graph Neural Networks.
Conclusion
The World, as all reality, everything and all that was, is and will be, is modeled, encoded or mapped for machine intelligence and learning and and human intelligence as a Multi-level Globally Distributed Causal Network, which is formally represented as the Global Giant Hypergraph (GGH).
SUPPLEMENT: CAUSAL DECALOGUE
The World Model Principles, Axioms and Assumptions
The world is the whole of reality, all and everything that was, is and will be, the totality of entities and interrelations.
The universe consists of substances (objects), states (qualities and quantities) and changes (actions, evets, processes) standing in causal interrelationships.
Causation is the master principle and prime force of the universe.
There are no uncaused things or changes in the world.
Causality gives structure or order to everything in the world, from the microworld to the macroworld.
Causation determines the hierarchical structure of world, its entities, processes and relationships, as well as its data, information and knowledge.
Causation is reverting, reversing, reacting, and going backwards, creating dynamic backward loops and causal circuits, complex control systems and nonlinear processes.
Causation flows in bottom-up ways, from micro to macro scales, as causal emergency, and vice versa, in top-down ways, from macro to macro scales, as causal control.
The interrelationships of microscales and macroscale are determined by the top-down and bottom-up causation.
Causality is the universal interrelationship, which is a symmetric productive correlative relationship, X causes Y if and only if Y causes X.
Corollaries
The world is the totality of all entities, structured by nonlinear causation, or interactions.
The world is the global causal net. All is causal networks.
The physical world or universe is the totality of all matter and energy, space and time; all that is, has been, and will be, all governed by causal interactions, as fundamental forces, gravitational, electromagnetic, strong and weak.
Causation gives deep structures and ordering to mind, intelligence, learning, inference, cognition. reasoning, understanding, and action, human or machine.
Causal models, rules and relationships are the master models and algorithms for artificial intelligence, machine learning, artificial neural networks and deep learning.
General conclusions about the world
Everything is interconnected to everything else.
Everything in the world is interconnected.
All is interrelated with all.
Sometimes the causal links are obvious, and most times undiscovered, unmeasurable or untraceable.
But the interconnections are there, and often affect us in ways we may never know.
The future is always harking back as a feedback and reaction, creating nonlinear complex structures, systems and processes.
All nonlinear dynamic systems, hurricanes, financial crises, animal stampedes, or the development of cities and civilizations, are examples of processes driven by nonlinear causality, as a self-reinforcing cause and effect.
Linear causality of cause and effect is just a part of nonlinear causality, which is a causal interaction or interactive causality.
The universe consists of objects having various qualities and standing in various relations, Whitehead and Russel
The world is a relational net; a mereological unity, a mereological whole embracing every part and constituent of the world, the world is a world of states of affairs, Armstrong.
The world is a world of facts, the world is a world of things, etc.
Resources
The CAUSAL DECALOGUE: The 10 Commandments about the world for humans and AI machines
--
4moHi sir I am teenagerI am exploring subjects I just want to ask you We see discipline or subjects in our school and universitiesLike science , commerce ,humanities etc They are division of human knowledge I want to know actually how many fields of study are there in total How many divisions of human knowledge we can do on very broadest level All disciplines that are studied in all universities of world Division of totality of known human knowledge Please respond 🙏