Preprint
Machine Learning

Artificial Intelligence and Human Enhancement: Can AI Technologies Make Us More (Artificially) Intelligent?

Sven Nyholm
August 30, 2023Cambridge Quarterly of Healthcare Ethics29 citations
DOI

29

Citations

0

Influential Citations

Cambridge Quarterly of Healthcare Ethics

Venue

2023

Year

Abstract

Abstract This paper discusses two opposing views about the relation between artificial intelligence (AI) and human intelligence: on the one hand, a worry that heavy reliance on AI technologies might make people less intelligent and, on the other, a hope that AI technologies might serve as a form of cognitive enhancement. The worry relates to the notion that if we hand over too many intelligence-requiring tasks to AI technologies, we might end up with fewer opportunities to train our own intelligence. Concerning AI as a potential form of cognitive enhancement, the paper explores two possibilities: (1) AI as extending—and thereby enhancing—people’s minds, and (2) AI as enabling people to behave in artificially intelligent ways. That is, using AI technologies might enable people to behave as if they have been cognitively enhanced. The paper considers such enhancements both on the level of individuals and on the level of groups.

Analysis

Why This Paper Matters

This paper addresses a central tension in the AI era: whether reliance on AI makes us less intelligent or can serve as a form of cognitive enhancement. As AI systems become more integrated into daily tasks, the question of their impact on human cognition is increasingly urgent. The paper provides a philosophical framework to navigate this debate, moving beyond simple optimism or pessimism.

By distinguishing between two forms of enhancement—extending the mind and enabling artificially intelligent behavior—the author clarifies the conceptual landscape. This is significant for AI practitioners and ethicists who need to evaluate the design and deployment of AI systems with cognitive effects in mind.

Technical Contributions

The paper's key contribution is a conceptual taxonomy of AI's potential cognitive effects. It introduces the idea of 'artificially intelligent behavior' as a distinct form of enhancement, where AI enables users to perform tasks as if they were more intelligent, without necessarily improving their internal cognitive capacities. This distinction is crucial for understanding the nature of AI-mediated cognition.

Additionally, the paper extends the discussion to group-level enhancement, considering how AI might enhance collective intelligence. This broadens the scope of cognitive enhancement beyond individual users.

Results

As a philosophical paper, there are no empirical metrics. However, the author's arguments are structured around two main possibilities: AI as mind extension (drawing on extended mind theory) and AI as enabling artificially intelligent behavior. The paper does not provide quantitative results but offers a qualitative analysis of these concepts.

Significance

This paper contributes to the growing field of AI ethics and philosophy of AI. It provides a nuanced perspective that can guide future research on human-AI interaction and cognitive enhancement. For AI practitioners, it highlights the importance of considering the cognitive impact of AI systems, not just their functional performance. The conceptual distinctions introduced could inform the design of AI tools that genuinely enhance human cognition rather than merely substitute for it.

artificial-intelligencehuman-enhancementcognitive-enhancementphilosophyethicshuman-intelligence

Related Papers