Amazing “Jailbreak” Bypasses ChatGPT’s Ethics Safeguards
OpenAI has been scrambling to enact new rules that prevent its wildly popular ChatGPT from generating text from being generally horrible — like by promoting things that are unethical, illegal, or just plain controversial. Back in the day, you could ask the red-hot AI system to whip up instructions on everything from shoplifting to committing Read more about Amazing “Jailbreak” Bypasses ChatGPT’s Ethics Safeguards[…]