Do large language models show decision heuristics similar to humans? A case study using GPT-3.5

J Exp Psychol Gen. 2024 Apr;153(4):1066-1075. doi: 10.1037/xge0001547. Epub 2024 Feb 8.


A Large Language Model (LLM) is an artificial intelligence system trained on vast amounts of natural language data, enabling it to generate human-like responses to written or spoken language input. Generative Pre-Trained Transformer (GPT)-3.5 is an example of an LLM that supports a conversational agent called ChatGPT. In this work, we used a series of novel prompts to determine whether ChatGPT shows heuristics and other context-sensitive responses. We also tested the same prompts on human participants. Across four studies, we found that ChatGPT was influenced by random anchors in making estimates (anchoring, Study 1); it judged the likelihood of two events occurring together to be higher than the likelihood of either event occurring alone, and it was influenced by anecdotal information (representativeness and availability heuristic, Study 2); it found an item to be more efficacious when its features were presented positively rather than negatively-even though both presentations contained statistically equivalent information (framing effect, Study 3); and it valued an owned item more than a newly found item even though the two items were objectively identical (endowment effect, Study 4). In each study, human participants showed similar effects. Heuristics and context-sensitive responses in humans are thought to be driven by cognitive and affective processes such as loss aversion and effort reduction. The fact that an LLM-which lacks these processes-also shows such responses invites consideration of the possibility that language is sufficiently rich to carry these effects and may play a role in generating these effects in humans. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

MeSH terms

  • Affect
  • Artificial Intelligence*
  • Communication
  • Heuristics*
  • Humans
  • Language