Related Concept Videos
Stereotype Content Model
Natural and Artificial Concepts
Components of Language
Language Development
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
Language and Cognition
Automatic Processing and Automatic Social Behavior
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Directionality and representativeness are differentiable components of stereotypes in large language models.
ChatGPT as Research Scientist: Probing GPT's capabilities as a Research Librarian, Research Ethicist, Data Generator, and Data Predictor.
REFORMS: Consensus-based Recommendations for Machine-learning-based Science.
Extracting intersectional stereotypes from embeddings: Developing and validating the Flexible Intersectional Stereotype Extraction procedure.
Erratum for the Research Article "Detecting supramolecular organic nanoparticles during heat wave".
Related Experiment Video
Updated: Jun 14, 2026

Experimental Paradigm for Measuring the Effect of Induced Emotion on Grammar Learning
Published on: January 29, 2020
Semantics derived automatically from language corpora contain human-like biases.
Aylin Caliskan1, Joanna J Bryson1,2, Arvind Narayanan1
1Center for Information Technology Policy, Princeton University, Princeton, NJ, USA. aylinc@princeton.edu jjb@alum.mit.edu arvindn@cs.princeton.edu.
Machine learning models trained on web text replicate human semantic biases found in Implicit Association Tests. This reveals how historical biases are embedded in language data, offering methods to identify and address cultural biases in technology.
Area of Science:
- Artificial Intelligence
- Natural Language Processing
- Computational Social Science
Background:
- Machine learning (ML) derives artificial intelligence by identifying patterns in data.
- Human language corpora contain implicit societal biases.
- The Implicit Association Test (IAT) measures the strength of automatic associations between concepts.
Purpose of the Study:
- To investigate whether machine learning models trained on human language exhibit human-like semantic biases.
- To determine if ML models can replicate biases measured by the IAT.
- To explore the potential of ML for identifying and mitigating cultural biases.
Main Methods:
- Applied a statistical machine learning model to a large text corpus from the World Wide Web.
- Trained the model on standard text data.
- Evaluated the model's semantic associations against known human biases, including those measured by the IAT.
Main Results:
- The ML model replicated a spectrum of human semantic biases, mirroring IAT results.
- Biases were observed across various domains, including morally neutral (insects, flowers), problematic (race, gender), and veridical (gender and careers/names).
- Text corpora accurately imprint historical human biases, which are recoverable through ML analysis.
Conclusions:
- Machine learning models trained on real-world text data inherit and reflect human semantic biases.
- Text data serves as a repository for historical biases, which can be quantified using ML.
- The developed methods offer a promising approach for detecting and addressing cultural and technological biases.
