Click here to view original web page at Red Teaming Improved GPT-4. Violet Teaming Goes Even Further
Last year, I was asked to break GPT-4—to get it to output terrible things. I and other interdisciplinary researchers were given advance access and attempted to prompt GPT-4 to show biases, generate hateful propaganda , and even take deceptive actions in order to help OpenAI understand the risks it […]
