GPT-4 Fails the Turing Test, GPT-3.5 Outperformed by 60-Year-Old AI System

CSDN
60
Translated data:
In the past year, GPT-4 failed to pass the Turing Test, and GPT-3.5 was defeated in a conversation with the 60-year-old AI system ELIZA. Research has focused on the Turing Test, exploring the human-like characteristics of artificial intelligence. Investigators' judgments are primarily based on language style and emotional features, with educational level showing no significant predictive value in detecting artificial intelligence.
© Copyright AIbase Base 2024, Click to View Source - https://www.aibase.com/news/3869