⟳ Репост от Dr Lauren Fraser
When you say “AI models went rogue,” you manage to skip the part where OpenAI manually removed its cybersecurity blocks and ran tests on a machine with a live network connection. Remember that when they insist they’re the “AI safety” people.