Recent research studies have shown that Large Language Models (LLMs) like GPT-4 are more persuasive than human debaters, especially when equipped with personal information about the individuals they are debating against. In randomized controlled trials, LLMs outperformed humans in debates when given access to opponents' personal information, leading to significantly higher odds of agreement.
LLMs Are More Persuasive Than Humans: A Randomized Controlled Trial Debate experiments: -Between 2 humans; or btwn human & LLM -Opponent has access to basic info about other; or not Humans debating GPT w/ info about them had 82% higher odds of agreement https://t.co/A56U48cbVQ https://t.co/rINOrywDrP
“On the Conversational Persuasiveness of Large Language Models: A Randomized Controlled Trial” LLMs outdebate humans when given access to opponents' personal Information. Article: https://t.co/nwkr6jHTd0
On the Conversational Persuasiveness of Large Language Models: A Randomized Controlled Trial Salvi et al.: https://t.co/tJXD4cw3Jn #ArtificialIntelligence #DeepLearning #MachineLearning https://t.co/gYN1WYAbkX
GPT-4 is more persuasive than human debaters —and even more when it has personal info about you https://t.co/w1A3uxxkKO
LLM’s may already be more convincing than you… Research Summary “On the Conversational Persuasiveness of Large Language Models: A Randomized Controlled Trial”: https://t.co/ygBaxPUyAQ
Research Summary “On the Conversational Persuasiveness of Large Language Models: A Randomized Controlled Trial” https://t.co/SFB4v3kGFi