Reproduce “Dissociating Artificial Intelligence from Artificial Consciousness” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Graham Findlay, William Marshall, Larissa Albantakis, Isaac David, William GP Mayner, Christof Koch, Giulio Tononi

This research provides a theory-driven investigation into artificial consciousness, demonstrating that functional equivalence in artificial intelligence does not imply phenomenal equivalence. Using Integrated Information Theory (IIT), the study shows that a digital computer perfectly simulating another system's behavior does not replicate its consciousness; instead, the computer's intrinsic causal structure typically results in fragmented or trivial conscious states.

Key insights

  • Computational functionalism, which posits that consciousness arises from performing the 'right kind' of computations, is fundamentally challenged as functional equivalence does not guarantee phenomenal equivalence.
  • The intrinsic causal structure of a system, as measured by Integrated Information Theory (IIT), determines its consciousness, not merely its external functions or observable behavior.
  • Artificial intelligence systems can achieve high levels of intelligence without necessarily possessing consciousness, highlighting a potential dissociation between these two properties in engineered systems.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.