A: Push back on a correct answer and watch the model fold.
B: Change one word in a famous riddle and watch a very capable model get it wrong.
C: Is AI a bubble, or the biggest build out in tech history?
D: GPT sounds like a black box.
E: Ask an AI to summarise a YouTube video and it often cannot.
F: You trained a model, tested it once, and the score looked great.
G: Your model scores brilliantly in training and falls apart on real data.
H: Give a model a score to chase, and it will chase the score.
I: A model writes one word at a time, and each word waits for the one before it.
J: Being listed in Anthropic's directory is not a security audit.
K: Same number of messages as last month, much bigger bill.
