Why Large Language Models Appear to Be Intelligent and Creative: Because They Generate Bullsh*t!
- 23 May 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 58 (6) , 91-96
Abstract
This article tries to explain why so many perceive large language models (LLMs), such as ChatGPT, as intelligent and creative. An LLM generates convincingly cogent bullsh*t (BS) in the Frankfurtian sense. This BS is perceived by humans as intelligent and creative.Keywords
Funding Information
- NSERC (RGPIN-2023-03584)
This publication has 10 references indexed in Scilit:
- The Promethean Dilemma of AI at the Intersection of Hallucination and CreativityCommunications of the ACM, 2024
- Correction: ChatGPT is bullshitEthics and Information Technology, 2024
- Generative Artificial Intelligence, Semantic Entropy, and the Big SortComputer, 2024
- Fatal Flaws in ChatAI as a Content GeneratorComputer, 2023
- ChatGPT's Astonishing Fabrications About Percy LudgateIEEE Annals of the History of Computing, 2023
- ChatGPT and AIChat EpistemologyComputer, 2023
- Bullshit Ability as an Honest Signal of IntelligenceEvolutionary Psychology, 2021
- The Oxford Handbook of Creativity, Innovation, and Entrepreneurship2015
- On Bullshit2005
- Reply to G. A. Cohen2002