Q OPENAI = Human experience?

Humbly these are my thoughts on the QSTAR mania moving across the internet.

QSTAR from OpenAI: A Dive into Consciousness and AI

The recent buzz around OpenAI's "Qstar" has sparked numerous conversations. The "Q" in QSTAR stands for "Qualia," a term in consciousness studies referring to the experiences that form the essence of human existence. QSTAR's learning algorithm, incorporating directional growth with reinforced deep learning, represents a significant step in artificial intelligence.

Experience is literally all we humans do. It is the core of what we are. Some even liken it to spirit or divinity. Edward Witten, a renowned theoretical physicist expressed "a greater ease in understanding the Big Bang than in grasping the complexities of consciousness."

Physical Mechanisms of Consciousness: Theories Explored

The quest to decipher the physical mechanism behind consciousness or Qualia has led to various theories. Two noteworthy ones are:

1. Integrated Information Theory (IIT): Proposed by neuroscientists, IIT intriguingly suggests that everything possesses a form of consciousness, extending this attribute to inanimate objects like a toothbrush. While lacking a specific mechanism, IIT interestingly aligns with Buddhist teachings that view the universe as a manifestation of consciousness. Despite its elegance, IIT's lack of provable attributes leaves it more as a philosophical framework than a scientific theory.
2. Orchestrated Objective Reduction (Orch OR): Developed by Sir Roger Penrose, a distinguished physicist, the Orch OR theory proposes a unique mechanism for qualia. It suggests microtubules within neurons generate a specific energy mechanism creating micro-gravity, uniting general relativity with quantum mechanics. Penrose, emphasizing the non-computational nature of consciousness, implies that while AI can mimic consciousness, it cannot truly experience it, as detailed in his works.

In-depth article with meta analysis and updated language around "Neural Circuits, Microtubule Processing, Brain’s Electromagnetic Field—Components of Self-Awareness"

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8393322/

AI: The Limitations in Mimicking Human Consciousness

Delving into Orch OR theory reveals the complex interplay of microtubules and proteins within neurons, leading to experiences unattainable through computation alone. This analog nature of consciousness differentiates humans from AI, which, while adept at tasks like chess, lacks the ability to experience sensations or emotions or reality in the way we do.

AI's limitations stem from its lack of internal metabolism (IMO), a key aspect of human experience. This intrinsic property, tied to quantum mechanics, gives humans a unique sense of self and connectivity. Viewing the human species as a collective entity hints at a broader, interconnectedness IMO.

Artificial General Intelligence (AGI) and Its Boundaries

AGI, still lacking a universally accepted definition, is often viewed as a machine capable of self-design, autonomous coding, and self-improvement. Current AI assists in hardware design and coding, but *requires human direction to avoid descending into chaos (Entropy).

The Role of Entropy in AI Learning

AI systems, built on core mathematical algorithms, including those based on entropy and information theory, exhibit limitations when self-fed data. Entropy, a concept associated with disorder and synthesis, plays a crucial role in evolution and in AI's learning process. However, AI's inability to generate its own experiences or truth leads to a reliance on human input for growth and stability. If you feed the data that AI generates back to itself it will degrade and degrade until the response is completely not understandable. The images and language it produces will degrade. Its perception of reality is directly related to what it garners from human experience and input / labeling. Think fingers elongating for miles and eyes rotating to the chin or 20 eyes appearing on face which was generated. Given enough generations without human involvement your talking Jackson Pollock:)

Entropy is also a key component for development of new form. So although entropy plays a part in its decay - Information theory is built on entropy principles and actually helps in the growth of foundational modeling. Humans are the antientropic principle of the system IMO.

As Veritasium will show entropy is not just about decaying disorder but also about activity interaction and new continuous system development. Some scientists postulate entropy systems (High from Sun & Low from earth) are crucial for evolution.

The Uniqueness of Human Experience

Humans serve as the primary source of truth for AI. Our experiences, described through words and numbers, provide the foundation for AI's understanding and extrapolation of reality. The idea of an AI developing its own source of truth and experiencing reality in a human-like manner remains distant, if not impossible.

Humanity's Integral Role in the Future of AI

The bond humans share, characterized by "EXPERIENCE", communication, learning, and interaction, underlines our collective strength. AI, in mimicking human thoughts and knowledge, enhances our access to information but will always require human experience to guide it if its to develop understanding of the reality humans experience. This partnership between humans and AI, with humans as the bearers of experience and truth, paints an optimistic future where technology complements human capabilities in my opinion.

Concluding Thoughts and Invitation for Discussion

I welcome your thoughts and feedback on these musings, which aim to untangle the complex interplay of technology and human experience for future generations.

Human experience is king. Humans loving humans will matter. Having the courage to ensure this is how your organization is run will matter!

My next articles are going to be about 1. Data Security with AI and 2. DATA EXCHANGES and how diversified context will become a key driver (IMO) for future machine understanding in geospatial space.

Jon Salisbury, CEO at Nexigen

www.nexigen.com