Intelligence Requires Grounding But Not Embodiment

TL;DR

This paper challenges the view that embodiment is necessary for intelligence by arguing that grounding suffices to support four core properties of intelligence: motivation, predictive ability, understanding of causality, and learning from experience. It shows how grounding can be realized in non-embodied digital environments and outlines a thought experiment with a tool-augmented language model operating online to demonstrate autonomous, goal-directed behavior without physical embodiment. Grounding provides a mechanism to attach external referents to symbols, enabling value assignment, causal learning, and experiential optimization, while prediction can be accomplished through self-supervised statistics. The authors thereby reframes the AI design problem, suggesting grounding in digital environments as a minimal prerequisite for intelligent behavior with broad implications for safety, scalability, and system architecture.

Abstract

Recent advances in LLMs have reignited scientific debate over whether embodiment is necessary for intelligence. We present the argument that intelligence requires grounding, a phenomenon entailed by embodiment, but not embodiment itself. We define intelligence as the possession of four properties -- motivation, predictive ability, understanding of causality, and learning from experience -- and argue that each can be achieved by a non-embodied, grounded agent. We use this to conclude that grounding, not embodiment, is necessary for intelligence. We then present a thought experiment of an intelligent LLM agent in a digital environment and address potential counterarguments.