Scientists shocked to find AI’s social desirability bias “exceeds typical human standards”
A new study published in PNAS Nexus reveals that large language models, which are advanced artificial intelligence systems, demonstrate a tendency to present themselves in a favorable light when taking…