Slightly off topic, but since the HN site and API is open to all, it'd be silly to assume our comments aren't also part of several datasets used to train LLMs.