AI2030 Responsible AI Tools Marketplace

Key Takeaways

Value alignment is not a new problem, but AI makes it uniquely urgent: While technology has long been capable of "locking in" particular value systems, the paper argues that AI systems' increased power and autonomy create unprecedented challenges in ensuring they reflect human values rather than embedding problematic ones.
Fairness and safety communities should work together: The paper draws important connections between work on algorithmic fairness, accountability, and transparency (often called "FAT") and technical AI safety research, suggesting these fields have more in common than typically recognized and should collaborate more closely.
The real challenge is "social value alignment," not just individual values: Rather than aligning AI to a single set of values, the paper emphasizes the need to address "social value alignment"—how to align AI systems with the diverse and sometimes conflicting values held by different groups of people, particularly at a global scale.
Participatory design offers a proven model: The paper references existing work on aligning technologies with social values through participatory design processes, suggesting these established approaches could inform how we tackle AI value alignment challenges.

Abstract

This paper addresses the question of how to align AI systems with human values and situates it within a wider body of thought regarding technology and value. Far from existing in a vacuum, there has long been an interest in the ability of technology to 'lock-in' different value systems. There has also been considerable thought about how to align technologies with specific social values, including through participatory design-processes. In this paper we look more closely at the question of AI value alignment and suggest that the power and autonomy of AI systems gives rise to opportunities and challenges in the domain of value that have not been encountered before. Drawing important continuities between the work of the fairness, accountability, transparency and ethics community, and work being done by technical AI safety researchers, we suggest that more attention needs to be paid to the question of 'social value alignment' - that is, how to align AI systems with the plurality of values endorsed by groups of people, especially on the global level.

Key Takeaways

Value alignment is not a new problem, but AI makes it uniquely urgent: While technology has long been capable of "locking in" particular value systems, the paper argues that AI systems' increased power and autonomy create unprecedented challenges in ensuring they reflect human values rather than embedding problematic ones.
Fairness and safety communities should work together: The paper draws important connections between work on algorithmic fairness, accountability, and transparency (often called "FAT") and technical AI safety research, suggesting these fields have more in common than typically recognized and should collaborate more closely.
The real challenge is "social value alignment," not just individual values: Rather than aligning AI to a single set of values, the paper emphasizes the need to address "social value alignment"—how to align AI systems with the diverse and sometimes conflicting values held by different groups of people, particularly at a global scale.
Participatory design offers a proven model: The paper references existing work on aligning technologies with social values through participatory design processes, suggesting these established approaches could inform how we tackle AI value alignment challenges.

Abstract

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

Key Takeaways

Abstract

Paper Info

Responsible AI Topics

Links

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

Key Takeaways

Abstract

Paper Info

Responsible AI Topics

Links