The Challenge of Value Alignment: from Fairer Algorithms to AI Safety

📖 Description

Forthcoming in The Oxford Handbook of Digital Ethics The Challenge of Value Alignment: from Fairer Algorithms to AI Safety Iason Gabriel and Vafa Ghazavi 1. Introduction There has long been a view among observers of artificial intelligence (AI) research, often expressed in science fiction, that it poses a distinctive moral challenge. This idea has been articulated in a number of ways, ranging from the notion that AI might take a ?treacherous turn? and act in ways that are opposed to the interests of its human operators, to deeper questions about the impact of AI on our understanding of human identity, emotions and relationships. Yet over the past decade, with the growth of more powerful and socially embedded AI technologies, discussion of these questions has become increasingly mainstream. Among topics that have commanded the most serious attention, the challenge of ?value alignment? stands out. It centres upon the question of how to ensure that AI systems are properly aligned...

📊 Game Impacts

💭 Reactions

🔗 Sources

• [1] https://arxiv.org/abs/2101.06060