Making AI Smarter for Humans

The prevailing conception of artificial general intelligence (AGI) often centers on single-agent multimodal large language models (LLMs) and physical models that mimic human linguistic and physical intelligence. However, this perspective overlooks the importance of decentralized, interacting systems in producing generalizable, adaptive behavior.

Drawing inspiration from statistical mechanics, neuroscience, and agent-based modeling, we argue that AGI is more appropriately conceptualized as an emergent property of interacting intelligent agents. These multi-agent systems, analogous to particles in thermodynamic ensembles or neuronal assemblies in the brain, possess the capacity for robust coordination, planning, adaptability, and resilience - properties absent in standalone models.

We propose a theoretical framework rooted in entropy and emergent dynamics and argue that AGI will not emerge from scale alone but from structured interactions among specialized systems.

Large language models (LLMs) have enabled powerful advances in natural language understanding and generation. Yet their application to complex, real-world scientific workflows remain limited by challenges in memory, planning, and tool integration.

Here, we introduce SciBORG (Scientific Bespoke Artificial Intelligence Agents Optimized for Research Goals), a modular agentic framework that allows LLM-based agents to autonomously plan, reason, and achieve robust and reliable domain-specific task execution. Agents are constructed dynamically from source code documentation and augmented with finite-state automata (FSA) memory, enabling persistent state tracking and context-aware decision-making.

This approach eliminates the need for manual prompt engineering and allows for robust, scalable deployment across diverse applications via maintaining context across extended workflows and to recover from tool or execution failures. We validate SciBORG through integration with both physical and virtual hardware, such as microwave synthesizers for executing user-specified reactions, with context-aware decision making and demonstrate its use in autonomous multi-step bioassay retrieval from the PubChem database utilizing multi-step planning, reasoning, agent-to-agent communication and coordination for execution of exploratory tasks.

Systematic benchmarking shows that SciBORG agents achieve reliable execution, adaptive planning, and interpretable state transitions. Our results show that memory and state awareness are critical enablers of agentic planning and reliability, offering a generalizable foundation for deploying AI agents in complex environments.