Yesterday, I gave the AI Dungeon version of the model the questions from the first round of the most recent episode of Jeopardy (almost certainly not in its tra

Yesterday, I gave the AI Dungeon version of the model the questions from the first round of the most recent episode of Jeopardy (almost certainly not in its training set). After I burned the first category on priming it, so it knew the kind of task we were doing, it got 72% of the remaining questions right. That's not just making it look like it knows what it's being asked.