Thinking Without Phenomenal Character: How are Artificial Intelligence, Cognitive Phenomenology, and Extended Cognition Related?

This paper is intended to provide a rough overview of the relationship between artificial intelligence (AI) and cognitive phenomenology in addition to extended cognition. Both cognitive states and phenomenal states are possible states had by humans, objects, systems, or other entities, but these different kinds of states are distinguishable. There is little to no reason to insist that machines be phenomenally conscious to be thinking, intelligent creatures on their own, or be phenomenally conscious to play a role in intelligent extended cognitive systems. Many varieties of cognitive phenomenology suggest that thinking requires conscious experience, which entails machines must be phenomenally conscious to think or to be intelligent. Positive views about cognitive phenomenology unnecessarily complicate the relationship between cognition and phenomenal character, and the very nature of cognitive states. Extended accounts of cognition also suggest that cognitive states and phenomenal states are distinguishable and may bear no necessary connection. As I will suggest, it is possible and ontologically simpler to explain thought without appealing to proprietary or any other phenomenal character. On this ontologically simpler view of cognition, a machine can theoretically be a thinking thing without being in a phenomenal state.