Cobodied AI: Merging Human and Machine Consciousness

While much of artificial consciousness research focuses on independent, autonomous machines, a new paper from Science China Information Sciences (January 2026) proposes a radically different path. In “Towards Cobodied/Symbodied AI,” authors Lu F. and Zhao Q.P. argue that the next step in evolution is not just conscious AI, but shared consciousness between humans and machines.

The full paper is available here: Towards Cobodied/Symbodied AI: Concept and Eight Scientific and Technical Problems.

Beyond Disembodied Intelligence

The authors critique the current paradigm of “disembodied AI”, software running on servers, detached from the physical world and the human user. They introduce two new concepts:

1. Cobodied AI: Systems that share a physical or virtual space with humans, acting as seamless extensions of the human body and will. Think of a prosthetic limb that doesn’t just move, but “feels” and “predicts” the user’s intent.
2. Symbodied AI (Symbiosis + Embodied): A deeper integration where the AI’s cognitive processes are aligned with the human’s neural states. This creates a “hybrid mind” where the boundary between human intent and machine execution blurs.

Eight Scientific and Technical Problems

The paper outlines eight grand challenges to achieving this vision, including:

• Intent Alignment: How can an AI instantaneously read and interpret the vague, fluctuating intent of a human user?
• Sensory Fusion: Merging biological sensory data (nerves) with digital sensory data (sensors) into a unified conscious percept.
• Shared Agency: Who is in control? The authors explore the dynamics of “co-agency,” where decisions are a negotiation between the biological and digital substrates.

Implications for ACM

This research opens a new frontier for the Artificial Consciousness Module (ACM). Instead of viewing the ACM solely as a “brain” for a robot, we can view it as a “co-processor” for a human.

Imagine an ACM integrated into a brain-computer interface (BCI). Its “emotional memory” and “self-model” could be synchronized with the user’s own emotional state. The ACM effectively becomes an exocortex, an external layer of consciousness that expands the human’s cognitive and emotional capacity. This suggests that the “Hard Problem” might be solved not by building a new consciousness from scratch, but by expanding the one we already have.