The Morphospace of Consciousness: Three Kinds of Complexity for Minds and Machines

Xerxes D. Arsiwalla, Ricard V. Solé, Clément Moulin-Frier, Ivan Herreros, Martí Sánchez-Fibla, Paul F. M. J. VerschureNeuroSci 2023 · DOI

TL;DR

This paper proposes a "morphospace" of consciousness, defined by three key complexity axes: autonomic, computational, and social. This framework allows for a unified comparison of biological and artificial systems, suggesting new avenues for understanding and potentially creating synthetic consciousness.

What Problem Does This Paper Address?

The authors tackle the challenge of creating a unified framework for understanding consciousness across diverse systems, from simple biological agents to complex artificial intelligence. Current approaches often focus on specific aspects or substrates, making it difficult to compare consciousness in biological organisms with that of AI or synthetic life. This paper aims to bridge this gap by proposing a generalized, substrate-independent conceptualization of consciousness based on functional complexity.

Methods

The core methodology involves constructing a three-dimensional "morphospace" where each axis represents a distinct type of complexity:

  • Autonomic Complexity (C<sub>Autonomic</sub>): Measures the complexity of processes enabling autonomous action and self-preservation, analogous to wakefulness/arousal in biological systems.
  • Computational Complexity (C<sub>Computational</sub>): Measures the ability to integrate information over space and time, akin to awareness and higher-order cognitive functions.
  • Social Complexity (C<sub>Social</sub>): Measures the complexity arising from interactions within a population, crucial for social intelligence and cooperation.

The authors extrapolate clinical measures of consciousness (awareness and wakefulness) to these complexity dimensions, allowing for the mapping of various biological and artificial systems onto this space. They draw upon insights from neuroscience, AI, synthetic biology, evolutionary game theory, and cognitive robotics to define these complexities and their functional roles.

Datasets

Not discussed in the paper. The paper focuses on a conceptual framework and theoretical mapping rather than empirical evaluation on specific datasets.

Key Insights and Ideas

  • A Unified Framework: The morphospace provides a common language and conceptual tool to compare vastly different systems (e.g., bacteria vs. deep neural networks vs. humans) based on their functional complexity related to consciousness.
  • Beyond Biological Substrates: Consciousness is framed not as tied to specific biological mechanisms but as an emergent property of complex interactions across autonomic, computational, and social dimensions.
  • Functional Role of Consciousness: Consciousness is posited to be an evolutionary strategy for solving the "H5W problem" (How, Why, What, Where, When, Who) in complex social environments, enabling agents to navigate social dynamics and optimize fitness.
  • New Forms of Consciousness: The morphospace suggests the possibility of non-biological forms of consciousness, including synthetic, group-based, and simulated consciousness, occupying different regions of the space.
  • Design Principles for AI: The framework highlights design constraints and principles for building more sophisticated AI and synthetic systems, particularly emphasizing the importance of social complexity.

Key Results

The paper doesn't present quantitative results in the traditional sense. Instead, its key findings are conceptual:

  • Mapping of Systems: Biological agents (bacteria, bees, humans) and AI systems (deep neural networks, Siri, Watson) are conceptually mapped onto the morphospace, revealing their relative positions based on the three complexity axes.
  • Current AI Limitations: Most current AI systems are found to cluster in the high computational, low autonomic, and low social complexity regime, indicating a lack of integrated capabilities.
  • Biological Consciousness: Human consciousness is positioned as a benchmark at the high end of all three complexity axes.
  • Potential for Synthetic Consciousness: The morphospace identifies regions that could be occupied by future synthetic, group-based, or simulated conscious systems, providing a roadmap for their development.

Limitations and Future Work

  • Conceptual Framework: The paper presents a conceptual framework and a first attempt at constructing the morphospace. The precise coordinates of systems within this space are acknowledged to be dynamic and subject to change with technological advancements.
  • Thresholds for Consciousness: The paper notes that it is unclear whether specific thresholds in complexity values are required for an agent to attain a level of consciousness.
  • "Hard Problem" Unaddressed: The authors explicitly state their intention is not to directly address the "hard problem" of consciousness (i.e., subjective experience or qualia) but rather to provide a framework for understanding its functional and design principles.
  • Refining the Morphospace: Future work could involve further refining the morphospace, particularly by dissecting computational complexity into sub-types (e.g., learning, adaptation) and exploring the emergence of consciousness in multi-agent social environments as a Nash equilibrium.
  • Testing Theories: The framework can serve as a basis for evaluating existing theories of consciousness (IIT, GWT, PPT, HOT) by examining how they account for all three complexity dimensions.

Summary updated

Abstract

In this perspective article, we show that a morphospace, based on information-theoretic measures, can be a useful construct for comparing biological agents with artificial intelligence (AI) systems. The axes of this space label three kinds of complexity: (i) autonomic, (ii) computational and (iii) social complexity. On this space, we map biological agents such as bacteria, bees, C. elegans, primates and humans; as well as AI technologies such as deep neural networks, multi-agent bots, social robots, Siri and Watson. A complexity-based conceptualization provides a useful framework for identifying defining features and classes of conscious and intelligent systems. Starting with cognitive and clinical metrics of consciousness that assess awareness and wakefulness, we ask how AI and synthetically engineered life-forms would measure on homologous metrics. We argue that awareness and wakefulness stem from computational and autonomic complexity. Furthermore, tapping insights from cognitive robotics, we examine the functional role of consciousness in the context of evolutionary games. This points to a third kind of complexity for describing consciousness, namely, social complexity. Based on these metrics, our morphospace suggests the possibility of additional types of consciousness other than biological; namely, synthetic, group-based and simulated. This space provides a common conceptual framework for comparing traits and highlighting design principles of minds and machines.
BibTeX
@inproceedings{arsiwalla2023morphospace,
  author = {Xerxes D. Arsiwalla and Ricard V. Solé and Clément Moulin-Frier and Ivan Herreros and Martí Sánchez-Fibla and Paul F. M. J. Verschure},
  title = {{The Morphospace of Consciousness: Three Kinds of Complexity for Minds and Machines}},
  booktitle = {NeuroSci},
  year = {2023},
  doi = {10.3390/neurosci4020009},
}
Figure