Trends in the AI Revolution: Demystifying Consciousness?

Vijayan K. Asari, PhD, Professor in Electrical and Computer Engineering, Ohio Research Scholars Endowed Chair in Wide Area Surveillance, Director, University of Dayton Vision Lab (Center of Excellence), University of Dayton

Vijayan K. Asari, PhD, Professor in Electrical and Computer Engineering, Ohio Research Scholars Endowed Chair in Wide Area Surveillance, Director, University of Dayton Vision Lab (Center of Excellence), University of Dayton

Human intelligence can be defined as a fusion of the extensive source of memory accumulated through impressions from prior experiences and humans' extraordinary ability of imagination that evolves from the reflections of the stored memory. It is a mental ability to learn from experience, extract abstract concepts, adapt to new environments, extrapolate knowledge, and apply it in various situations. Advanced AI can precisely perform properly trained activities such as recognizing images, identifying objects, tracking objects, translating languages, generating images/videos from texts, creating texts from images/videos, operating robotic functionalities, predicting climatic changes, diagnosing medical abnormalities, and so on. It is making monumental strides in all walks of life, including healthcare, disease detection and prevention, therapeutics, finance and banking, agriculture, management, industry and commerce, manufacturing, transportation, construction, labor safety, environmental monitoring, and smart vehicles.

These AI systems, termed Artificial Narrow Intelligence (ANI), show outstanding performance in specific applications. However, they lack the exclusively complex cognitive abilities of human beings.

Artificial General Intelligence (AGI), or Strong AI, is a hypothetical type of intelligence system intended to accomplish any intellectual task that human beings can perform or beyond. AGI aims to create machines that can think, reason, learn, and adapt to new situations and environments like humans. It can generalize the learned information, simultaneously perform a large suite of complex tasks, and even extend to a self-learning machine that can automatically learn and understand data. AGI systems incorporate emotional intelligence that is able to perceive the emotional state of the human mind, such as tone, mood, sentiments, and fatigue, and control the functionalities of AI-based supporting systems accordingly. An AGI-based personal assistant could be very intelligent, intuitive, and interactive with humor, sarcasm, and other subtle clues in sharing emotions so that one can have conversations with another human being. AGI-based self-navigating systems could communicate with each other and adapt to real-life situations considering the weather, traffic, and other anomalies. Even though a powerful AGI system could involve human ingenuity, interpretability, and a better understanding of the factuality of the world and ethical considerations in coordinating with humans in real-life activities, it lacks the human’s capability of wild imaginations to the next phases based on the learned experiences and expertise.

The mystic theory defines human intelligence as a fusion of intellect, memory, identity, and consciousness. Intellect represents a human’s piercingly sharp capability to discriminate and distinguish sensory inputs. Memory stores the accumulated information from past experiences. Identity is the capability to categorically represent the distinguishable signatures in conjunction with the stored memory. Consciousness is a boundless evolution of the human mind beyond the existence of memory. Mystic science describes consciousness as memory-less intelligence beyond physical reality. Studies conducted in neuroscience observe the communication of information between cortices inside the brain. There are rhythmic oscillations of brain waves within individual cortices without intercortical communications during an unconscious state. There is an intercortical coordination and collaboration of brain waves in a conscious state of mind. Mystic science interprets this process of intercortical relays as a state of wakefulness and not necessarily a state of consciousness.

These theories hypothesize a practical definition of human consciousness as a hetero-associative generative phenomenon of the brain with interactive stimulations from multiple cortical regions. Each cortical processing subsystem receives reflections from other cortices and performs a hetero-associative generation of newer reflections. A fusion of the new reflections from multiple cortices produces the necessary signals to ignite the motor cortical functions to perform conscious activities. A hyperdimensional multimodal data analysis and multidirectional information fusion strategy is a feasible way of conceptualizing the fundamental structure of the model of consciousness in AI systems. This led to the development of Artificial Superintelligence (ASI), machine intelligence that could be much smarter than human brains in practically every application. ASI technology will be able to outperform human beings in social skills and scientific capability as its capabilities could extend towards a limitless cosmic space.

It is envisioned that in the future, a panel of AI systems will be able to perform as advisors to the heads of institutions, organizations, companies, businesses, corporations, or even countries in providing advisory support in strategic matters and decision-making. A properly trained AI expert system will be able to consider every aspect of rules, regulations, procedures, and strategies, and it could interpret the impacts without the possibility of missing any salient information from learned data. The human associates could reevaluate and assess the results and confirm the decisions as appropriate. The AI system will be able to incorporate any additional insights that the human advisers found to be necessary by its self-learning capability.

Like any other technological revolution, AI advancements must be auditable and traceable and should be constrained to strict internal and international laws, rules, and regulations for proper development and usage. The installation of diligent mechanisms to assess the impacts and oversights that avoid violations of human rights and threats to environmental well-being is necessary to secure public trust in the advancement of AI systems. A strong government-regulated, regulated, transparent, and explainable AI code of ethics should be incorporated to avoid bias, ensure the privacy of users and their data, and mitigate environmental and societal risks.