The Fundamentals of Reality: Principle of Universal Interaction: Real AI
Real intelligence, natural or machine, is the power to know, predict, and interact with the world, mentally or computationally simulating all possible predictions and effects, outcomes and consequences.
Or, to be intelligent means to know the world in all its reality and complexity.
To know reality is to know its fundamentals (foundations, essentials, first principles and rules, bedrock, ABC) for the deep understanding, efficient and sustainable interactions with the world at all its forms and levels.
Today's machine learning mostly covers one type of modeling, predictive modeling or predictive data analytics, while Real or True, Techno-Scientific AI involves all scientific modeling, simulating reality in all its fundamentality, complexity and interactivity. [Trans-AI: How to Build True AI or Real Machine Intelligence and Learning]
The Fundamentals of Reality
Universal Definitions.
Universal interaction is the fundamental and universal principle that everything in the universe interacts having an effect on each other
The world, W, is the universe of entities, E, and interactions, I: W = <E, I>
Reality as a whole, W = R, is the universal interactive network of things or entities
Universal Axioms.
The universal and fundamental concept of universal interaction depends on five axioms:
Types and Levels of Reality
The interactive reality (R) is the world of interactions at all its levels, from the quantum reality to the cosmological universe. It includes as its interacting parts all possible worlds or realities or environments:
The Real-World, Objective/Physical/Material Reality
The Mental World, Subjective/Personal/Mental Reality
The Social World, Intersubjective/Social Reality
The Technological World, Technological Reality
The Digital World, Digital Reality Environments
Simulated/Virtual Realities
Augmented/Enhanced Reality
Mixed/Extended Reality...
A special case of the digital IR is the metaverse, "an immersive, interactive environment generated by a computer". As the digital technologies that support the metaverse reality of immersive computing go:
Virtual reality (VR) headsets; Haptic interfaces; Intelligent sensors; Brain-computer interfaces (BCIs); Cryptocurrencies; Blockchain distributed ledgers; Nonfungible tokens (NFTs); Holograms; Digital twins; Augmented reality
Universal Interaction Laws
Universal self-causation is universal interaction. All change in the universe is a result of interactions. "Nowhere in the world can there be any phenomena that do not give rise to certain consequences and have not been caused by other phenomena".
The Law of Universal Causation from J.S. Mill:
Every phenomenon has a cause, which it invariably follows; and from this are derived other invariable sequences among the successive stages of the same effect, as well as between the effects resulting from causes which invariably succeed one another
The Law of Infinite Causes and Effects
Causation is an infinite chain of causes and effects, an infinite regress of causes and infinite progression of effects. If all effects are the result of previous causes, then the cause of a given effect must itself be the effect of a previous cause, which itself is the effect of a previous cause, and so on, forming an infinite logical chain of events that can have no beginning.
Any action, event, activity or change has numberless causes and effects, known or unknown, predicted or unpredicted, intended or unintended, anticipated or unanticipated, beneficial or harmful, good or bad...
The Law of Infinite Consequences refers to all actions, physical or social, including the development and commercialization of technologies, machines and structures. It is clear that many technologies bring huge benefits for humans and much harm for nature.
The same rule applies to digital technologies, social media, online gaming, mobile phones, ML tools, or LLMs.
In sociology, there is Merton’s Law of?Unintended Consequences to describe the unwelcome side effects of social actions, including technological innovations, explaining why and how commercial digital technologies fail consumers. Merton concluded that individuals would fail to comprehend all the outcomes arising from innovations out of ignorance, human error, or inexperience.
Real Intelligence is the power to know, predict, and interact with the world, ideally forecasting all major results and effects, outcomes and consequences.
It was the lack of predictive intelligence when the fossil fuel transportation was massively introduced, instead of e-cars at the beginning of the 20th century.
It is the want of predictive intelligence when machine rote learning tools and deepfake statistical software impersonated as AI, real machine intelligence and learning...
We need to think WHOLISTICALLY before doing something... trying to PREDICT all externalities.
Laws of Science and Technology: Ceteris Paribus (CP) Assumptions
Most laws and rules of science and technology are ceteris paribus laws and rules when "all other things being equal", "other things held constant", "all else unchanged", and "all else being equal".
A statement about a?causal,?empirical, or?logical?relation between two states of affairs is?ceteris paribus?if it is acknowledged that the statement can fail because of, or the relation can be abolished by, intervening factors.
领英推荐
A?ceteris paribus (CP)?assumption?limits key scientific research and experimental inquiry or technological inventions, because scientists or engineers seek to eliminate factors that perturb a relation or effect of interest.
Thus?there comes dependent?and?independent variables, as input and output, exposure and outcome,?in?mathematical modeling,?statistical modeling?and?experimental sciences.
Dependent endogenous variables depend, by some law or rule (e.g., by a mathematical function), on the values of other variables, like a multivariable regression analysis. Independent exogenous variables, in turn, are not depending on any other variable in the limited scope of the study.
Depending on the context, an independent causative variable is called a "predictor variable", "regressor", "covariate", "manipulated variable", "explanatory variable", "exposure variable", "risk factor", "feature" (in?machine learning?and?pattern recognition) or "input variable".
Then a dependent or caused variable is called a "response variable", "regressand", "criterion", "predicted variable", "measured variable", "explained variable", "experimental variable", "responding variable", "outcome variable", "output variable", "target" or "label".
A CP variable is a "controlled variable", "control variable", or "fixed variable", and confounders (confounding variables, confounding factors, extraneous determinants or lurking variables).
It could cover "extraneous variables", as subject variables, blocking/experimental or situational variables, as well as what is missed by the independent variable, uncorrelated with the regressors, the error variable, e, the "residual", "side effect", "error", "unexplained share", "residual variable", "disturbance", or "tolerance".
The CP variables are threats to the internal and external validity or transferability of studied causal inferences, the effect sizes between the variables, or generalizations of that inference to other contexts, the extent to which the results of a study can be replicated and generalized to and across other situations, people, stimuli, times, locations, etc.
The CP assumptions result in statistical associations or spurious relationships or spurious correlations, mathematical relationships?in which two or more events or variables are?associated?but?not?causally related, neither has a causal effect on the other.
Researchers or experiment designers seek to control?independent variables?as factors that may influence?dependent variables—the outcomes of interest or the hypothesis under study.?Omitted-variable bias (OVB) occurs when a statistical model leaves out one or more relevant variables.
In reality, it is not known which variable changed first, it can be difficult to determine which variable is the cause and which is the effect.
Scientific Models as CP modeling
A scientific model seeks to represent?empirical?objects, phenomena, and physical processes in a?restricted?way. All models are?in simulacra, that is, simplified reflections of reality that are poor approximations, sometimes useful.
So, scientific modelling?is "an activity that produces?models?representing?empirical?objects, phenomena, and physical processes, to make a particular part or feature of the world easier to?understand,?define,?quantify,?visualize, or?simulate. It requires selecting and identifying relevant aspects of a situation in the real world and then developing a model to replicate a system with those features. Different types of models may be used for different purposes, such as?conceptual models?to better understand, operational models to?operationalize,?mathematical models?to quantify,?computational models?to simulate, and?graphical models?to visualize the subject".
In?scientific modeling, simplifying assumptions permit illustration of concepts considered relevant to the inquiry of the selected or isolated cause-and-effect relationship.
"All models are wrong, but some are useful"
There is the common aphorism: "All models are wrong, but some are useful"; for all known scientific models are under CP assumptions:
Real AI Modeling and Simulation: RAI M & S
Today's machine learning mostly covers one type of modeling, predictive modeling or predictive analytics, while Real or True, Techno-Scientific AI involves all scientific modeling and simulating reality in all its fundamentality, complexity and interactivity.
There are generally three types of scientific experiments, in vitro, in vivo, and in silico. In silico experiments are carried out by computers from massive quantities of data, using powerful processors whose cores are made of silicon; in vivo experiments, performed on living matter, in vitro experiments, carried out in glass test-tubes.
Real AI involves the world model learning and inference and interaction engine, M & S and Computational Science, also known as?scientific computing,?technical computing?or?scientific computation?(SC), considered as a third mode of science complementing theory and observation/experimentation.
Computational science uses advanced?computing?capabilities to understand and solve any computable complex problems, as in:
Algorithms?(numerical?and non-numerical):?mathematical models,?computational models, and?computer simulations?developed to solve?sciences?(physical,?biological, and?social),?engineering, and?humanities?problems
Computer hardware?that develops and optimizes the advanced system?hardware,?firmware,?networking, and?data management?components needed to solve computationally demanding problems
The computing infrastructure that supports both the science and engineering problem solving and the developmental computer and?information science
AI S & C & M&S is using?the fundamental models of reality underlying simulated models?(e.g.,?physical, mathematical, behavioral, or?logical?representation of a?system, entity, phenomenon, or process)?to develop insightful knowledge for decision making and interactions.
RAI's development structure looks as:
Real/True/Trans AI = Fundamentals of Reality + World Model Learning, Inference and Interaction Engine + M & S + SC + AI Models + ML + DL algorithms + LLMs + Automation + Robotics + the Internet of Things + Digital Cloning + Extended Reality
Resources