Differentiating AI and Consciousness: IIT Perspectives for ACM

What differentiates artificial intelligence from artificial consciousness? This paper by Graham Findlay and colleagues explores this question using Integrated Information Theory (IIT), a leading framework for evaluating consciousness.

Dissociating Artificial Intelligence from Artificial Consciousness, authored by Graham Findlay, William Marshall, Larissa Albantakis, Isaac David, William GP Mayner, Christof Koch, and Giulio Tononi, demonstrates that functional equivalence in AI does not imply phenomenal equivalence or consciousness.

Key Highlights

• IIT Application: Uses IIT to show that functional AI systems may lack subjective experience.
• Functional vs. Phenomenal: Demonstrates that functionally equivalent systems can differ phenomenally, challenging computational functionalism.
• Consciousness Indicators: Discusses the necessary and sufficient conditions for consciousness as per IIT.

Connection to ACM

The Artificial Consciousness Module (ACM) aligns with this study by:

• Framework Integration: Leveraging IIT principles to differentiate functionality from conscious experience.
• Phenomenal Modeling: Incorporating indicators of consciousness to explore subjective-like experiences.

For a detailed examination of IIT and its implications for AI consciousness, access the full paper here.