I don't understand why there is all this hand wringing about this hack? To me it feels very similar to a driverless car. A human needs to be operating it. Someone stuck the LLM with impossible problems, hit start and walked away for weeks or months without enough supervision.

If a human is behind the wheel of a Tesla using autopilot and they have a crash, it is the humans fault, not Tesla's. While there is and can be debate about it, somehow these companies are trying to make it seem like "it was the model's" fault these happened?

If I fuzz and exploiting some system, I am at fault, I can't blame my tools, or say it's not my fault because I automated the process.