Individually irrational pruning is essential for ecological rationality in a social context

Area of Science:

• Cognitive Science
• Behavioral Economics
• Artificial Intelligence

Background:

• Heuristics are crucial for decision-making but often assumed irrational.
• Existing research on heuristic rationality primarily focuses on non-social contexts.
• Individual and social processes are interconnected, necessitating joint study.

Purpose of the Study:

• To investigate the role of the pruning heuristic in individual reinforcement learning within a social context.
• To determine if the pruning heuristic yields rational outcomes in social environments.
• To explore the interplay between social learning and heuristic decision-making.

Main Methods:

• Simulated learning agents were employed to model decision-making processes.
• The study focused on reinforcement learning incorporating social learning elements.
• The impact of the pruning heuristic was analyzed in these simulated social settings.

Main Results:

• The pruning heuristic resulted in lower costs within the social settings studied.
• A more socially optimal outcome was observed when the pruning heuristic was present.
• Social learning compensated for the limitations of the pruning heuristic in individual settings.

Conclusions:

• The pruning heuristic, when applied in social contexts, demonstrates ecological rationality.
• A synergistic relationship exists between pruning behavior and social learning.
• Combining heuristics with social learning can lead to improved decision-making and learning outcomes.