Intelligence and Consciousness

Abstract

One can trace a possible leap from intelligence to consciousness in the debates in and around the field of AI. We argue that this leap is a result of an unscrutinized assumption. We attempt to analyze the relationship between intelligence and consciousness by proposing a stripped-down working definition of intelligence. Based on this definition, we argue that there is already at least one system that is more intelligent than at least one of the conscious systems, yet that more intelligent system is not conscious. Therefore, there is not enough evidence to sustainably support the leap from intelligence to consciousness.

Keywords:

• Artificial intelligence
• consciousness
• strong AI

Disclosure Statement

On behalf of all the authors, the corresponding author states that there is no conflict of interest.

Notes

1 In an attempt to define machine intelligence, Legg and Hutter (2007) provide a comprehensive survey and analysis of a number of both formal and informal definitions provided for human intelligence and then attempt to extract essential features from these definitions to get hold of the general idea of intelligence. Their main aim in the paper remains to provide a formal definition of machine intelligence. Therefore, they formalize the essential features of intelligence that they extract mathematically to develop a general measure of intelligence for arbitrary machines.

2 In Dretske’s proposed definition of intelligence, any system that exhibits an intelligent behavior must have an underlying thought explaining the behavior. This thought underlying an intelligent behavior according to Dretske may or may not be a result of reasoning process or may or may not be true. The thought behind the intelligent behavior is a necessary condition but not a sufficient one. In other terms, thought alone is not enough for intelligence however it must be there. But why this connection between thought and behavior is necessary for intelligence is not clear. As a result, the definition of intelligence proposed by Dretske, even in the money sense, is necessarily tied to biological beings. In Dretske’s definition, all hardwired intelligent systems are excluded from the category of intelligence as the origin of genuine intelligence can be understood in terms of learning process. However, his proposal of learning process can be objected to on the basis of the following two grounds. One, if the role of learning process can be replaced by a hard-wired functionally equivalent feature, then the resultant intelligent behavior can be indistinguishable. Second, modern-day interactive AI systems based on language processing do learn in their interaction with human users and therefore alter their subsequent interactions based on what they have learnt. This goes against the general proposal of Dretske for intelligence as there is no thought involved in the processing of these systems.