Clever but not wise: The intelligence in generative AI

Description

Borrowing from the work of Daniel Dennet (1987), Mike argues that generative AI can be understood as genuinely intelligent, but in a fundamentally different way from human intelligence. Where the latter is personified by a deliberative capacity to select some ends over others, generative AI is equipped to find means to ends, where the ends themselves are exogenously specified (either by a manufacturer or by a user). Drawing on remarks from Philippa Foot (1977), Mike claims that this distinction broadly maps to differences between ‘cleverness’ and ‘wisdom’. AI might be extremely clever, but it is not wise. This might have implications for how AI should be used in education, and what skills students should be looking to develop in a world that is likely to be saturated with AI.

This is the fifth of a series of events from ARU’s AI Collaborations: a positive approach to the use of generative artificial intelligence in higher education.