Bringing AI to Autonomous Systems — From Cognition to Collective Intelligence (arXiv:2609.30291v1)
This new arXiv preprint presents a comprehensive framework for designing and evaluating autonomous systems, centering on a generic agent architecture that composes cognitive functions around a long-term memory of evolving knowledge. The paper argues for combining connectionist and symbolic approaches, integrating AI with systems engineering, highlights implementation challenges (sensory-to-structured-data links, goal-driven decision-making, multi-agent coordination), and proposes directions for assessing agent trustworthiness while noting a large gap between the vision and current capabilities.