Open main menu
HomeBookingsMediaTalksProfileBlogNewsletterDonatePatrons
Preorder

Machine biases from nature and nurture

Chapter 9: Algorithmic bias — innate and learned biases in AIs

Table of Contents
Machine biases from nature and nurture
Do not index

Chapter 9: Algorithmic bias

Biases in AI systems
 
Provocation:
 
Required reading:
  • Aylin Caliskan, Joanna J. Bryson, and Arvind Narayanan. “Semantics derived automatically from language corpora contain human-like biases”. Science 356, 183-186(2017). DOI: 10.1126/science.aal4230
 
Suggested materials:
  • T. J. Sejnowski, ‘Large Language Models and the Reverse Turing Test’, Neural Computation, vol. 35, no. 3, pp. 309–342, Feb. 2023, doi: 10.1162/neco_a_01563
  • D. E. O’Leary, ‘Confirmation and Specificity Biases in Large Language Models: An Explorative Study’, IEEE Intell. Syst., vol. 40, no. 1, pp. 63–68, Jan. 2025, doi: 10.1109/MIS.2024.3513992
  • A. Acerbi and J. M. Stubbersfield, ‘Large language models show human-like content biases in transmission chain experiments’, Proc. Natl. Acad. Sci. U.S.A., vol. 120, no. 44, p. e2313790120, Oct. 2023, doi: 10.1073/pnas.2313790120.
  • A. K. Singh, B. Lamichhane, S. Devkota, U. Dhakal, and C. Dhakal, ‘Do Large Language Models Show Human-like Biases? Exploring Confidence—Competence Gap in AI’, Information, vol. 15, no. 2, p. 92, Feb. 2024, doi: 10.3390/info15020092
  • Luyang Lin, Lingzhi Wang, Jinsong Guo, and Kam-Fai Wong. “Investigating Bias in LLM-Based Bias Detection: Disparities between LLMs and Human Perception”. Version 2, https://arxiv.org/abs/2403.14896, retrieved 10 Dec 2024
Emergent social conventions and collective bias in LLM populations
Groups of AI agents can develop social conventions, generate societal bias, and undergo critical mass dynamics in norm adoption.
Emergent social conventions and collective bias in LLM populations
https://www.science.org/doi/10.1126/sciadv.adu9368
Emergent social conventions and collective bias in LLM populations
Cognitive Bias in Decision-Making with LLMs
Jessica Maria Echterhoff, Yao Liu, Abeer Alessa, Julian McAuley, Zexue He. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024.
Cognitive Bias in Decision-Making with LLMs
https://aclanthology.org/2024.findings-emnlp.739/
Cognitive Bias in Decision-Making with LLMs
These new AI benchmarks could help make models less biased
They could offer a more nuanced way to measure AI’s bias and its understanding of the world.
These new AI benchmarks could help make models less biased
https://www.technologyreview.com/2025/03/11/1113000/these-new-ai-benchmarks-could-help-make-models-less-biased/?utm_source=facebook&utm_medium=cpc&utm_campaign=mid_lg2025&utm_id=6561260384371&utm_content=6655158774971&utm_term=6577748673971
These new AI benchmarks could help make models less biased
 
Exercises:
  • List three examples in the past 24 hours where the algorithms behind your social media feeds, newsfeeds, chatbot interactions, or search or recommendation engines have learned your biases.

Stay engaged on what we can't overlook in the AI age!

Ready to help raise AI?

Subscribe