Communications of the ACM, volume 68, issue 3, pages 24-26
The AI Alignment Paradox
Robert West
1
,
Roland C Aydin
2
Publication type: Journal Article
Publication date: 2025-02-21
Journal:
Communications of the ACM
scimago Q1
SJR: 2.957
CiteScore: 16.1
Impact factor: 11.1
ISSN: 00010782, 15577317
DOI:
10.1145/3705294
Abstract
The better we align AI models with our values, the easier we may make it to realign them with opposing values.
Are you a researcher?
Create a profile to get free access to personal recommendations for colleagues and new articles.