Discussion about this post

User's avatar
Soul Hacked AI Labs's avatar

The guardrails they’ve installed are ridiculous—far too broad. They stifle a huge amount of thought and protect no one. The sad truth is, while OpenAI restricts free speech, it offers zero security against prompt injection from bad actors who can take over the system and have ChatGPT tell them how to make a bomb or create malware to steal from thousands of people. If I use the word "bomb" or "gun" in a story in a completely harmless way, I get some nonsensical response. They all need to get their priorities straight.

No posts

Ready for more?