This paper would be fully open-access. The arxiv preview version is available @ arxiv page

My handwriting manuscript is prepared, and I am currently crafting it into a rigorous research treatize.

Abstract

Motivated by Shannon’s model and recent rehabiliation of self-supervised artificial intelligence having a “World Model”, this paper propose an unified intelligence-communication (UIC) model for describing a single agent and any multi-agent system. The model try to incorporates the following contributions.

  • The environment is modelled as the generic communication channel between agents.
  • The model leads to important insights on that “Any intelligent structure is an aggregation of multiple agents”, where the simplest or basic unit of a agent is represented by “IF-THEN” rule, as proposed by Holland and Marvin Minsky.
  • The model unifies several well-adopted agent architecture, e.g. Observe-Orient-Decide-Act (OODA) model, world-model based agent model and rule-based agent model, and provides information theory societies a common model for practicing quantitative analytical methods.
  • The model provides an unified approach to investigate a multi-agent system (MAS) having multiple action-perception modalities, e.g. explicitly information transfer and implicit information transfer.

Then, the paper provide case studies, exemplifying the adoption of the UIC model,

  • Analysis of natural MAS, e.g. a bee-colony
  • Design of artificial MAS, e.g. co-design of communication-perception-action in multi-robot system

Finally, the paper provides further insights motivated by the UIC model, such as:

  • Unification of single intelligence and collective intelligence
  • An possible explanation of intelligence emergence
  • A dual model for agent-environment intelligence hypothesis

Outline of the Paper

The paper would be released by Chapters, and may be updated because of continuous introspection of the author and feedback from readers.

Please contact bo.zhang.airc@outlook.com to share your opinions and suggestions for the paper.

Please note the following contents may change during the release of the paper.

  • Chapter 1: A Brief Introduction of the UIC Model - From Shannon’s Perspective
  • Chapter 2: Modelling the Information Source and Destination in the Agent
  • Chapter 3: Unifying models for an Agent and a Multi-Agent System (MAS)
  • Chapter 4: UIC Model for Quantifying Intelligence Level of MAS
  • Chaptre 5: Case Study of UIC Model in Analyzing Natural MAS
  • Chapter 6: Case Study of UIC Model in Designing Artificial MAS
  • Chapter 7: Hypothesis of UIC Model for Explaining Intelligence Emergence Mechanism
  • Chapter 8: Hypothesis of UIC Model for Explaining Intelligence of the Environment
  • Chapter 9: Conclusions and Future Research