It certainly makes you wonder what it is that OpenAI has in GPT-4o that allows that service to (presumably) answer the query without safety concerns (unless in fact it *is* unsafe and OpenAI has done nothing about it.)

→ View original post on X — @tiernanraytech