The General Theory of General Intelligence: A Pragmatic Patternist Perspective (2103.15100v3)

Abstract: A multi-decade exploration into the theoretical foundations of artificial and natural general intelligence, which has been expressed in a series of books and papers and used to guide a series of practical and research-prototype software systems, is reviewed at a moderate level of detail. The review covers underlying philosophies (patternist philosophy of mind, foundational phenomenological and logical ontology), formalizations of the concept of intelligence, and a proposed high level architecture for AGI systems partly driven by these formalizations and philosophies. The implementation of specific cognitive processes such as logical reasoning, program learning, clustering and attention allocation in the context and language of this high level architecture is considered, as is the importance of a common (e.g. typed metagraph based) knowledge representation for enabling "cognitive synergy" between the various processes. The specifics of human-like cognitive architecture are presented as manifestations of these general principles, and key aspects of machine consciousness and machine ethics are also treated in this context. Lessons for practical implementation of advanced AGI in frameworks such as OpenCog Hyperon are briefly considered.

Overview of the General Theory of General Intelligence: A Pragmatic Patternist Perspective

Ben Goertzel's paper, "The General Theory of General Intelligence: A Pragmatic Patternist Perspective," provides a comprehensive overview of decades of exploration into the theoretical foundations of artificial and natural general intelligence (AGI). This work synthesizes core philosophies, formalizations, and proposed architectures that ground AGI in a coherent conceptual framework. The exploration is intended to bridge the critical gap between abstract theorization and practical system implementation.

Goertzel's work explores several key themes essential for understanding and engineering AGI systems:

Philosophical Foundations and Ontology

The core of the paper revolves around the "patternist philosophy of mind," which posits that intelligence is essentially a system's capability to recognize and manipulate complex patterns in itself and its environment. Central to this is the concept of cognitive synergy—multiple cognitive processes, each handling distinct types of knowledge and patterns, work together, augmenting each other's strengths and compensating for weaknesses.

To formalize the notions of pattern, Goertzel introduces a mathematical theory that intertwines algorithmic information theory with new constructs—such as distinction graphs and combination systems. These foundational constructs allow the formulation of complex cognitive structures, embedding the principles of pattern recognition and transformation.

Artificial General Intelligence Architectures

Part of the paper's ambition is a proposed high-level AGI architecture, which draws from the explored theoretical principles. At its core is the concept of discrete decision systems (DDSs) articulated through combinatory function optimization (COFO). Here, the framework integrates different cognitive apparatuses in a metagraph-based system, emphasizing a unified knowledge representation to enhance cognitive synergy. This metagraph approach serves as a sophisticated structure underpinning AGI's cognitive processes, allowing distinct cognitive algorithms to interoperate efficiently and evenly distribute computational resources.

Implications of Cognitive Design and Implementation

Goertzel suggests that the outlined AGI principles and architecture, specifically the OpenCog Hyperon system, embody these theoretical concepts, aiming for practical AGI realization. However, the pervasive challenge remains bridging the theoretical constructs with scalable, practical applications. The paper also addresses the necessity of AGI systems that not only perform goal-directed behavior but are adaptable and capable of self-modification to maintain ethical integrity and alignment with human values.

Consciousness and Ethics in AGI

Human-like consciousness and the ethical dimensions of AGI are addressed through a reflection on self and meta-learning—core to achieving AGI with reflective ethical qualities. By focusing on a holistic approach, Goertzel advocates for AGI systems that can engage in ethical reasoning and empathetic considerations, enhancing the potential for AGI systems that are beneficially integrated into the broader human context.

Continuing Challenges and Prospects

The paper does not claim the problem of AGI is solved; rather, it recognizes that while theoretical constructs have reached substantial depth, the transition to a powerful AGI remains a formidable path. Prospects for accelerating progress are promising, contingent on advances in computational and conceptual frameworks. Development and deployment of such systems hold significant potential for advancing AGI but are also bound by the profound ethical and societal responsibilities they entail.

In conclusion, Ben Goertzel’s exploration of a general theory of general intelligence provides a structured and thought-provoking framework incorporating philosophical, cognitive, and ethical perspectives. While detailed in its approach, the pathway from theory to fully realized AGI systems remains an ongoing pursuit, creating opportunities for prospective work and research directions grounded in these foundational insights.