Cognitive Carrying Capacity: Self-Representation in Energy-Constrained Agents

In this paper, we consider how Cognitive Carrying Capacity (CCC) determines the extent to which artificial agents model the self within resource-constrained environments. We approach the question from an AI perspective, asking what cognitive mechanisms enable artificial agents to generate self-reports similar to those of their biological counterparts. We argue that the key mechanisms include attention, environmental modeling, and memory, all operating under the central constraint of energy limitations. Attention filters environmental stimuli through internal goals/costs and restricts the range of environmental models the agent can generate. This energy limit leads to the narrowed correlation of goals/costs with specific models of self as separated from the rest of the environment. Unlike more traditional accounts, ours sees the self/world separation as a function of energy constraints, aligning with Dennett’s view that the self is not a fixed entity but a dynamically evolving construct (Dennett, 1991). The Cognitive Carrying Capacity of an agent establishes a threshold that defines the boundaries of the self as one of a group of models the agent constructs. The larger the energy resources, the larger and more complex the self-representation becomes. By adjusting energy constraints in computational experiments, we explore how CCC predicts the self-representation threshold in multi-agent environments, allowing for a dynamic and adjustable scale of self-representation, resulting in agents "seeing" themselves in some cases as a sole agent, while in others as the whole universe. We argue that CCC limitations is the reason biological agents cannot see themselves as the whole Universe.

1. Cognitive Carrying Capacity: A Dennettian Approach to self-Representation in Energy-Constrained Agents Nikos Tzagkarakis Keith Frankish

2. Why don’t we think of the self as the whole universe? Bear with me

3. The self as the universe Is there a particular reason the boundaries exist? Can the Self be a scale? Do environmental pressures play a role? Nikos Tzagkarakis, Keith Frankish

4. What is the self? (in our model) First things first

5. What is the self The self is not a singular entity—it is the efficiently inferred subset of the world most relevant to the agent's survival and goals The cluster of patterns in the environment that are most predictively entangled with the agent’s needs, goals, and risks. ➤ A functional boundary—not a metaphysical one. Nikos Tzagkarakis, Keith Frankish

6. The role of the environment.

7. Environmental Pressures Environmental complexity demands selective modeling—computation becomes currency. Agents operate under energy constraints. Each model costs computation. The more complex the environment, the more expensive self-modeling becomes. ➤ The self shrinks to fit what the agent can afford to model. Nikos Tzagkarakis, Keith Frankish

8. Cognitive Carrying Capacity A measure of how much self-related complexity an agent can sustain, given its energy and computational limits. Cognitive Carrying Capacity (CCC) =The maximum complexity of self-models an agent can maintain given its limited energy, attention, and memory resources. ➤ Determines the scale and boundary of the self. ➤ Directly shaped by environmental demands and survival needs. Nikos Tzagkarakis, Keith Frankish

9. The PhD Research.

10. Group Compositions of Foraging Agents under Environmental Constraints Nikos Tzagkarakis, Keith Frankish

11. What is a Foraging Agent Definition: Autonomous entities that search for, collect, and potentially return resources within an environment. Key components: Search strategy (random, systematic, informed) Resource detection capabilities Collection/carrying capacity Return mechanisms Energy consumption model And more… Foraging serves as a fundamental paradigm for studying resource allocation, collective behavior, and adaptive strategies. Nikos Tzagkarakis, Keith Frankish

12. internal states as a variable Our Focus

13. Research Question How does the scale of internal-state representation affect foraging outcomes under strict energy constraints? Investigating the relationship between: Agent cognitive complexity Energy efficiency Collective resource collection performance Nikos Tzagkarakis, Keith Frankish

14. Agent Types: The Key Distinction Type A: Self-Only Internal State Maintains only own internal information (goals, costs, etc.) Relies on immediate local sensory inputs Lower computational and energy demands No representation of other agents' states Type B: Enhanced Internal State Maintains models of self AND other agents Estimates peer goals and states Higher computational and energy requirements Potential for improved coordination without explicit communication

15. Methodology: Controlled Environment Current multi-agent foraging research has not addressed: Maze-like foraging environment with fixed resource patches Predetermined energy budgets Clearly defined navigation challenges Systematic variation of energy availability Key Question: How does the qualitative difference in internal state representation transform collective foraging performance under varying energy constraints? Nikolaos Tzagkarakis

16. Evaluation Metrics Current multi-agent foraging research has not addressed: Resource collection efficiency Resources gathered per unit of energy expended Overall Energy Consumption Total energy used over fixed simulation time Task Completion and Longevity Time to collect target resources / Operational duration without energy depletion Nikos Tzagkarakis, Keith Frankish

17. Testable Trade-Offs Type B agents may achieve better coordination in high-energy scenarios Type A agents may prove more efficient when energy is scarce Critical thresholds likely exist where the advantage shifts between types The value of internal state sharing is expected to vary with environmental complexity Environmental pressure may determine population dynamics Nikos Tzagkarakis, Keith Frankish

18. Contribution This research represents the first systematic investigation of: How the interchange of internal states between agents affects foraging performance The specific energy trade-offs of maintaining different internal representations The conditions where enhanced internal modeling justifies its energy costs Design principles for optimal agent architecture based on energy availability Nikos Tzagkarakis, Keith Frankish

19. Nikos Tzagkarakis [email protected] Thank You!