Hacker News new | past | comments | ask | show | jobs | submit login

Just tried it and ChatGPT’s 4o model got it wrong like you said until I added the line “think step by step”, which is always a good idea for these kinds of logic puzzles.

ChatGPT o1 got the answer correct with no tweaks to the prompt.




So it's great for answering questions you know the answer for and can guide it if necessary?




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: