Are Humans Evolved Instruction Followers? An Underlying Inductive Bias Enables Rapid Instructed Task Learning

2026-06-29

AI Summary

How do humans reliably execute novel tasks correctly on their first attempt based solely on linguistic instructions? This work proposes that this capacity arises from an inductive bias for instruction following, shaped by evolutionarily endowed cognitive architectures, which enables rapid behavioral generalization from language. Integrating theoretical and empirical insights from cognitive science, neuroscience, and machine learning, the study draws a cross-disciplinary analogy between this human ability and instruction fine-tuning in large language models. It formulates a unified theoretical framework that systematically characterizes instruction following as a shared mechanism for fast learning in both natural and artificial neural systems, and outlines testable, interdisciplinary research pathways to further investigate this phenomenon.

Abstract

Human adults can often perform a novel task correctly on the first attempt after only receiving verbal or written instructions. This rapid instructed task learning (RITL) is a hallmark of human cognitive flexibility, yet its mechanisms and parallels in artificial systems remain under-explored across disciplines. In this position paper, we argue that humans possess an evolved instruction-following bias -- an inductive bias shaped by evolution to interpret and execute linguistic instructions which critically enables fast generalization of behavior from language. This bias functions analogously to the way large language models (LLMs) leverage instruction tuning to achieve zero-shot task performance. We synthesize evidence from cognitive science, neuroscience, and machine learning research to support this hypothesis. While instruction-following in AI is currently achieved via specialized training protocols, we posit that in humans it arises as an innate cognitive architecture feature. We outline testable predictions and call for more interdisciplinary research to investigate Instruction-Following as a unifying mechanism enabling rapid task learning in both natural and artificial neural networks.

Problem

• rapid instructed task learning
• instruction-following
• inductive bias
• cognitive flexibility
• evolved bias

Innovation

• instruction-following bias
• rapid instructed task learning
• inductive bias
• cognitive architecture
• zero-shot generalization