Ontogenetic AGI Frameworks: Developmental Pathways toward General Machine Intelligence

R. K. Sharma, P. L. Verma, S. D. Pati, M. Joshi

Abstract


ABSTRACT Artificial General Intelligence (AGI) is expected to perform wide variety of intellectual tasks similar to human intelligence. Traditional AI systems mostly rely on large datasets and static training methods, but human intelligence grows gradually through developmental stages from infancy to adulthood. This concept of gradual learning is known as ontogeny in biological sciences. Ontogenetic AGI frameworks propose that machine intelligence should also follow developmental trajectory, where learning happens progressively through stages, guided by environment interaction, self-exploration, memory growth, and cognitive maturation. This paper reviews the concept of ontogenetic AGI, its theoretical foundations, cognitive inspirations, architectural designs, and practical implementations. Different developmental stages of machine intelligence, role of curiosity, embodiment, memory evolution, and learning scaffolds are discussed. The study also presents architectural diagrams and comparative tables to illustrate differences between static AI models and ontogenetic AGI frameworks. Finally, challenges and future research directions are highlighted for building self-developing intelligent systems.

KEYWORDS: AGI, Ontogeny, Developmental Learning, Cognitive Architecture, Self-learning Systems, Curiosity-driven Learning, Embodied Intelligence


Full Text:

PDF 161-174

Refbacks

  • There are currently no refbacks.