“Machine Intelligence, not Artificial Intelligence and What Really is Machine Learning?” describes a journey in an effort to reach an adaptive machine intelligence. Whereas humans seek to adapt and control this world’s environment, machines are designed and engineered by humans to control processor, memory and storage functions in the machine’s environment. Instincts are born with humans and automatically or in response to thought, humans desire and seek satisfaction, and build machines. Instincts are not born into the machine, there are no thoughts and there are no desires. Humans are adaptive and seek to control existence. Machines do not naturally seek or desire to control anything. The machine’s hardware provides superior information processing, management, and storage that greatly benefits humanity. Complementing the machine’s hardware, software solutions are installed or developed for useful purposes.
Are we satisfied with how solutions are currently provided to us, or do we want something more flexible? Although not self-aware, an additional layer that extends the machine’s adaptability with interfaces that humans or machines instruct is more flexible. Not reaching the instinctive nature of a human, this layer enables the machine with automatic response enabling the flexible layer. In hindsight, will we see this journey as evolving to simplicity that is much like the transition from moving parts to solid-state, and receiving significant benefit such as horse and buggy compared to jet airliner?
Data and behaviour are basic relative concerns that align human needs with the machine capability.
Common today, brute-force is required through software development and installation to make data and behavior useful. Brute-force often results in rigid data and behavior for specific purposes. Instances where software solutions do not satisfy need, other software solutions are searched, other software is developed, software is modified, and more brute-force is required to achieve the next state. When solutions adapt from instructed needs, brute-force is minimized. With a layer respecting data in the machine similar to genotype in concept, and respecting behavior similar to phenotype in concept, we can establish process to entity relationships as concept models. Relationship metadata and meta operators receive concept models as newcomers into the Supermodel. Archetypes then adapt by stepping up or down which drives innate ability. Understanding that automation must go beyond diagnostic result extending into long-term functional life-cycles is required and is the important differentiator. There is no need of a developer, architect, or other intermediary specialist as this approach allows the person that knows the business to automate the business by instructing concept models. Complexity is then reduced to interface(s) any authorized user or machine instructs. Eventually many solutions are hosted by the Supermodel for any users or machines to then use on-demand.
Conceptually, this is not new when considering that model-based reasoning combines model knowledge with observed data. Evolving model-based reasoning to a Supermodel is even more useful. A proven physical Supermodel now working in production at global scale for years is achieved combining Cybernetic concepts with Enterprise Architecture derivations to form archetypes.
“Cybernetics is a transdisciplinary approach for exploring regulatory systems—their structures, constraints, and possibilities. Norbert Wiener defined cybernetics in 1948 as “the scientific study of control and communication in the animal and the machine.” In other words, it is the scientific study of how humans, animals and machines control and communicate with each other.”
Whereas a human’s heartbeat, hearing, vision, breathing, and other functions are instinctive automatic response, the machine’s Supermodel has automatic response that is not a human mimicry but is instead a well-defined ecosystem of trans and inter disciplines fixed as systems within the Supermodel. Further disciplines are added to the Supermodel as needs demand. However, each set of variant instructions composed as concept models are distinct and different based on human desires to automate legal, insurance, regulatory, government or other automation as a service. Collectively, concept models are assimilated and unified into the Supermodel. Emphasis again on the difference being that the concept model forms for each unique automation need. Trans and interdisciplinary hidden layers acting as archetypes determine when and how to automatically materialize, engage and operate from concept models. Foundational context and operational context forms automatically, from the inside-out, as active metadata resulting from operations and is the next foundation for deeper automation with machine learning.
Think about the chicken and the egg, do we learn before we think or do we think before we learn? What is the underlying base that supports intelligence? One does not question how the heartbeats; we do not ask how the machine’s processor and memory work, and the Supermodel experience will eventually be the same.
Not unlike the human’s physical environment the cyber environment also has physical resources.
Whereas this world has resources humans adapt, the cyber world has resources the machine adapts. The human’s world is different than the cyber world and this difference should be easy to accept, but some seem to think that there is a secret essence that is beyond human consciousness. Why seek an artificial intelligence when a machine intelligence is achievable and simply assists man’s need to further seek and adapt? What is common between the human and machine is data and behavioral alignment. The human need is complemented by the machine’s capability. With the adaptive Supermodel reacting to concept models’, computer science evolves beyond rigid concrete boundaries and derived clusters. Instead, a flexible and fluid layer establishes context from the inside-out that adapts to the instructed human needs. Machine adaptation is realized by a series of synthetic operations similar to fixed action pattern. Automatic mappings occur as entities and control forms on-demand from concept models. Cybernetics recognizes differences between human, animal, and machine and the commonality in control theory, system theory, mathematical communications and other supporting disciplines. The need is for the machine to adapt to the human need. With the Supermodel, efficiencies are increased by orders of magnitude.
In simple terms, the Supermodel provides a machine-controlled, on-demand, layer where humans or machines instruct the Supermodel, no code, configurable, Automation as a Service, with genotype and phenotype properties automatically materialize as archetypes control solutions for other humans or machines to use.
1) Automation as a service (AaaS) market is projected to reach $10,914.9 million by 2023
2) Global Low-Code Development Platform Market is Set to Reach USD 53.0 billion by 2024, Observing a CAGR of 80.0% during 2019–2024
3) Case Management Market to Grow US$ 10 Billion by 2024
“Scarcely anyone who comprehends this theory can escape its magic.”
About the Author:
Richard Yawn is a technology leader, architect, developer, inventor and visionary with experience providing solutions across industries and functions in insurance, manufacturing, government and legal.