It is rather interesting to see that the safety mechanisms for #LLM amount to a couple of prompts hiding under a trench coat: https://arstechnica.com/information-technology/2023/02/ai-powered-bing-chat-spills-its-secrets-via-prompt-injection-attack/
@seanfobbe Interesting!
I tried that with ChatGPT,
results are in the picture below.
