Are AIs getting dangerously good at persuasion? OpenAI says “not yet.” - Ars Technica

Retrieved on: 2025-02-03 17:33:57

Tags for this article:

Click the tags to see associated articles and topics

Are AIs getting dangerously good at persuasion? OpenAI says “not yet.” - Ars Technica. View article details on hiswai:

Excerpt

OpenAI says its models are more persuasive than 82 percent of Reddit users. ChatGPT maker worries about AI becoming “a powerful weapon for ...

Article found on: arstechnica.com