Episode Details
Back to Episodes
A New Trick Uses AI to Jailbreak AI Models—Including GPT-4
Published 2 years, 10 months ago
Description
Adversarial algorithms can systematically probe large language models like OpenAI’s GPT-4 for weaknesses that can make them misbehave. Read the story here.
Learn about your ad choices: dovetail.prx.org/ad-choices