ChatGPT-4 Matches Humans in Personality and Decision-Making, New Study Finds
-
Researchers evaluated AI chatbots like ChatGPT against human personality and decision-making benchmarks.
-
ChatGPT-4 closely matched median human personality scores, while ChatGPT-3 differed somewhat.
-
In behavioral games, the chatbots showed human-like tendencies for fairness, cooperation, and learning from experience.
-
Framing and context significantly swayed the chatbots' strategies, indicating sophisticated responsiveness.
-
The study offers new benchmarks for assessing AI abilities to complement human traits and abilities.