Fun fact: When posts are published by first-time accounts, we submit them to LLMs with a prompt that asks it to evaluate whether they’re spammy, whether they look LLM-generated, etc, and show the result next to the post in the new-user moderation queue. The OpenAI API refused to look at this post, returning 400 Invalid prompt: your prompt was flagged as potentially violating our usage policy.
That is hilarious. I guess it’s not really surprising, because we pretty much discuss every maximally AI-guidelines taboo. Always appreciate a little look behind the scenes ;)
Fun fact: When posts are published by first-time accounts, we submit them to LLMs with a prompt that asks it to evaluate whether they’re spammy, whether they look LLM-generated, etc, and show the result next to the post in the new-user moderation queue. The OpenAI API refused to look at this post, returning
400 Invalid prompt: your prompt was flagged as potentially violating our usage policy.That is hilarious. I guess it’s not really surprising, because we pretty much discuss every maximally AI-guidelines taboo. Always appreciate a little look behind the scenes ;)
Now I’m wondering what the UI there looks like.