Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Case Repositories: Towards Case-Based Reasoning for AI Alignment (2311.10934v3)

Published 18 Nov 2023 in cs.AI, cs.CY, and cs.HC

Abstract: Case studies commonly form the pedagogical backbone in law, ethics, and many other domains that face complex and ambiguous societal questions informed by human values. Similar complexities and ambiguities arise when we consider how AI should be aligned in practice: when faced with vast quantities of diverse (and sometimes conflicting) values from different individuals and communities, with whose values is AI to align, and how should AI do so? We propose a complementary approach to constitutional AI alignment, grounded in ideas from case-based reasoning (CBR), that focuses on the construction of policies through judgments on a set of cases. We present a process to assemble such a case repository by: 1) gathering a set of ``seed'' cases -- questions one may ask an AI system -- in a particular domain, 2) eliciting domain-specific key dimensions for cases through workshops with domain experts, 3) using LLMs to generate variations of cases not seen in the wild, and 4) engaging with the public to judge and improve cases. We then discuss how such a case repository could assist in AI alignment, both through directly acting as precedents to ground acceptable behaviors, and as a medium for individuals and communities to engage in moral reasoning around AI.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (34)
  1. Case-based reasoning: Foundational issues, methodological variations, and system approaches. AI communications, 7(1):39–59, 1994.
  2. A general language assistant as a laboratory for alignment. arXiv preprint arXiv:2112.00861, 2021.
  3. Constitutional ai: Harmlessness from ai feedback. arXiv preprint arXiv:2212.08073, 2022.
  4. On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’21, page 610–623, New York, NY, USA, 2021. Association for Computing Machinery. ISBN 9781450383097. doi: 10.1145/3442188.3445922. URL https://doi.org/10.1145/3442188.3445922.
  5. The nature of the judicial process. Quid Pro Books, 2010.
  6. Case law grounding: Aligning judgments of humans and ai on socially-constructed concepts. arXiv preprint arXiv:2310.07019, 2023.
  7. Goldilocks: Consistent crowdsourced scalar annotations with relative uncertainty. Proceedings of the ACM on Human-Computer Interaction, 5(CSCW2):1–25, 2021.
  8. Case based reasoning as a model for cognitive artificial intelligence. In Case-Based Reasoning Research and Development: 26th International Conference, ICCBR 2018, Stockholm, Sweden, July 9-12, 2018, Proceedings 26, pages 62–77. Springer, 2018.
  9. Robin Downie. Health care ethics and casuistry. Journal of medical ethics, 18(2):61, 1992.
  10. Towards Measuring the Representation of Subjective Global Opinions in Language Models, 2023.
  11. Robert K Fullinwider. Philosophy, casuistry, and moral development. Theory and Research in Education, 8(2):173–185, 2010.
  12. Iason Gabriel. Artificial Intelligence, Values and Alignment. Minds and Machines, 30(3):411–437, Sep 2020. ISSN 0924-6495, 1572-8641. doi: 10.1007/s11023-020-09539-2. arXiv:2001.09768 [cs].
  13. Thomas C Grey. Langdell’s orthodoxy. U. Pitt. L. Rev., 45:1, 1983.
  14. Aligning ai with shared human values. arXiv preprint arXiv:2008.02275, 2020.
  15. Oliver Wendell Holmes Jr. The common law. Routledge, 2020.
  16. Albert R Jonsen. Casuistry and clinical ethics. Theoretical Medicine, 7:65–74, 1986.
  17. Robert A Kagan. Adversarial Legalism: The American Way of Law. Harvard University Press, 2019.
  18. Janet L Kolodner. An introduction to case-based reasoning. Artificial intelligence review, 6(1):3–34, 1992.
  19. A general introspective reasoning approach to web search for case adaptation. In International Conference on Case-Based Reasoning, pages 186–200. Springer, 2010.
  20. Quark: Controllable text generation with reinforced unlearning. Advances in neural information processing systems, 35:27591–27609, 2022.
  21. John Leslie Mackie. Hume’s moral theory. Routledge, 2003.
  22. A holistic approach to undesired content detection in the real world. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 15009–15018, 2023.
  23. Meta-case-based reasoning: self-improvement through self-understanding. Journal of Experimental & Theoretical Artificial Intelligence, 20(1):1–36, 2008.
  24. Training language models to follow instructions with human feedback. In S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh, editors, Advances in Neural Information Processing Systems, volume 35, pages 27730–27744. Curran Associates, Inc., 2022. URL https://proceedings.neurips.cc/paper_files/paper/2022/file/b1efde53be364a73914f58805a001731-Paper-Conference.pdf.
  25. Norbert Paulo. Casuistry as common law morality. Theoretical Medicine and Bioethics, 36(6):373–389, 2015.
  26. Whose opinions do language models reflect? arXiv preprint arXiv:2303.17548, 2023.
  27. Evidence-driven retrieval in textual cbr: bridging the gap between retrieval and reuse. In Case-Based Reasoning Research and Development: 23rd International Conference, ICCBR 2015, Frankfurt am Main, Germany, September 28-30, 2015. Proceedings 23, pages 351–365. Springer, 2015.
  28. Polis: Scaling deliberation by mapping high dimensional opinion spaces. Recerca: revista de pensament i anàlisi, 26(2), 2021.
  29. Adam Smith. The theory of moral sentiments, ed. dd raphael and al macfie, 1976. VII, iv, 34.
  30. Value kaleidoscope: Engaging ai with pluralistic human values, rights, and duties, 2023.
  31. Learning to summarize with human feedback. Advances in Neural Information Processing Systems, 33:3008–3021, 2020.
  32. Cass R Sunstein. Legal reasoning and political conflict. Oxford University Press, 2018.
  33. John Tasioulas. Artificial intelligence, humanistic ethics. Daedalus, 151(2):232–243, 2022.
  34. Fine-grained human feedback gives better rewards for language model training. arXiv preprint arXiv:2306.01693, 2023.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (5)
  1. K. J. Kevin Feng (13 papers)
  2. Quan Ze Chen (13 papers)
  3. Inyoung Cheong (7 papers)
  4. King Xia (2 papers)
  5. Amy X. Zhang (58 papers)
Citations (8)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now