OpenAI used the subreddit, r/ChangeMyView, to create a test for measuring the persuasive abilities of its AI reasoning models ...
OpenAI just released o3-mini, a reasoning model that’s faster, cheaper, and more accurate than its predecessor.