Toward Integrated AGI Ecosystems: A Review, Framework, and Research Agenda for Cyber-Physical-Social-Thinking Intelligence

Abstract

Artificial General Intelligence (AGI) represents a transformative frontier of machine intelligence, aiming for human-level adaptability across diverse domains. However, AGI research remains fragmented, with most advancements focusing narrowly on algorithmic, computational, or neuroinspired components. This review introduces Cyber-Physical-Social-Thinking(CPST) perspective to understand how AGI orchestrates complex multi-domain interactions, offering a unifying framework as an integrated intelligence system. Following PRISMA 2020 guidelines, we analyzed and 139 peer-reviewed studies(2021-October 2025), identifying three dominant research trajectories: (1) cognitive architecture integration, (2)social intelligence in distributed environments, and (3) reality-aware adaptation mechanisms. Our analysis identifies critical challenges in ethical governance, system resilience, and scalable learning across heterogeneous domains. We propose a forward-looking research agenda emphasizing hybrid human-machine intelligence, trustworthy AGI systems, and sustainable CPST ecosystems. This study advances intelligent systems research by providing an integrative framework that bridges AI algorithms with real-world deployment considerations, human factors, and societal impacts essential dimensions for next-generation intelligent technologies.

Ähnliche Arbeiten