People cannot distinguish GPT-4 from a human in a Turing test
arxiv:2405.08007
Retrieval, embeddings, benchmarks · 2024-05-09
Complete cleaned Markdown · Source Markdown including bibliography
Read in parts
Contents
- People cannot distinguish GPT-4 from a human in a Turing test
- Abstract
- 1 Introduction
- 1.1 The Turing test
- 1.2 AI models
- 1.3 Game setup
- 2 Results
- 2.1 Pass rates
- 2.2 Strategies & reasons
- 3 Discussion
- 3.1 Does GPT-4 pass the Turing test?
- 3.2 What does the Turing test measure?
- 3.3 Robots in disguise
- 4 Methods
- 4.1 AI witnesses
- 4.2 Interface
- 4.3 Participants
- 4.4 Procedure
- 4.5 Strategy and Reason Tagging
- 4.6 Statistical Analysis
- Acknowledgements
- Appendix A Full Prompt
- Appendix B Supplementary Figures
- Appendix C Reason and Strategy Classification Schemes
- Instructions for reporting errors