Coincidence or not, they seem to be poking fun at it: https://openai.com/index/learning-to-reason-with-llms/#chain...

(end of the cipher example)