Probing for Consciousness in Machines

Artificial agents trained via reinforcement learning can develop rudimentary forms of self and world models—key components of core consciousness as defined by Antonio Damasio. In a virtual environment, an agent learning to play a video game formed internal representations that allowed probes (feedforward classifiers) to predict the agent's spatial position from its neural activations. These results suggest that machine consciousness may be possible as a byproduct of goal-directed learning, offering foundational insights for AI development.

Study at a glance

Abstract

This study explores the potential for artificial agents to develop core consciousness, as proposed by Antonio Damasio's theory of consciousness. According to Damasio, the emergence of core consciousness relies on the integration of a self model, informed by representations of emotions and feelings, and a world model. We hypothesize that an artificial agent, trained via reinforcement learning (RL) in a virtual environment, can develop preliminary forms of these models as a byproduct of its primary task. The agent's main objective is to learn to play a video game and explore the environment. To evaluate the emergence of world and self models, we employ probes-feedforward classifiers that use the activations of the trained agent's neural networks to predict the spatial positions of the agent itself. Our results demonstrate that the agent can form rudimentary world and self models, suggesting a pathway toward developing machine consciousness. This research provides foundational insights into the capabilities of artificial agents in mirroring aspects of human consciousness, with implications for future advancements in artificial intelligence.