A matter of principle? AI alignment as the fair treatment of claims
Iason Gabriel & Geoff Keeling (2025)
One-Sentence Thesis
The authors propose an alternative account of AI alignment that focuses on fair processes, arguing that principles generated through these processes are the appropriate target for alignment.
Argument Outline
- 1Introduction to the normative challenge of AI alignment
- 2Critique of existing approaches to AI alignment (intent-alignment and HHH)
- 3Proposal of a novel approach to AI alignment based on fair processes
- 4Defense of the new approach and its advantages over existing approaches
Key Distinctions
Technical vs. normative aspects of AI alignment
Intent-alignment vs. HHH approaches to AI alignment
Fair process-based views vs. other approaches to AI alignment
Key Terms
AI alignment
Intent-alignment
HHH (Helpful, Honest, Harmless)
Fair process-based views
Flashcards
19 cardsRelated Questions
In Iason Gabriel & Geoff Keeling's "A matter of principle? AI alignment as the fair treatment of claims", Leike et al. supports which of the following?
In Iason Gabriel & Geoff Keeling's "A matter of principle? AI alignment as the fair treatment of claims", Gabriel criticizes which of the following?
Which of the following does Iason Gabriel & Geoff Keeling criticizes in "A matter of principle? AI alignment as the fair treatment of claims"?
In Iason Gabriel & Geoff Keeling's "A matter of principle? AI alignment as the fair treatment of claims", Crenshaw, K. defines which of the following?
In Iason Gabriel & Geoff Keeling's "A matter of principle? AI alignment as the fair treatment of claims", Rawls supports which of the following?
In Iason Gabriel & Geoff Keeling's "A matter of principle? AI alignment as the fair treatment of claims", Fair Process defines which of the following?