> Maybe Studio Ghibli making it through the seemingly deterministic GPT guardrails was an OpenAI slip up, a mistake,

The author is so generous... but Sam Altman literally has a Ghibli-fied Social profile and in response to all this said OpenAI chooses its demos very carefully. His primary concern is that Ghibli-fying prompts are over-consuming their GPU resources, degrading the service by preventing other ChatGPT tasks.

The official White House account has been posting ghiblified images too, Altman knows that as long as he's not critical of the current administration he's untouchable.