I’m really sorry that the bot didn’t perform as expected in this case. Thank you so much for trying it out and providing this detailed feedback—this is exactly the kind of input we need to improve!
There are a few reasons for this issue. The prototype can sometimes struggle with accessing its internal knowledge consistently, which leads to inaccuracies like the one you encountered. I’ve done my best to minimize this, but I’m still fully limited by the OpenAI platform. I also noticed in the screenshots that it “searched 3 sites” when it really shouldn’t have—it should rely entirely on its internal Jupiter knowledge. I’ve now disabled that feature. Additionally, I’m working on updating the prototype so it responds in the same language as the question.
Based on your feedback, I tested this specific question multiple times, and while the bot occasionally gets it right, it’s definitely not yet at the level we want. That said, I can confidently promise that the Beta will answer this question correctly 10 out of 10 times (or maybe 9.9 out of 10—you never know! 
). In fact, with the very small part of the Beta that we’ve already built, it’s getting the answer right every time (or at least I think so—my French isn’t that good, but I do see the number 9 coming up! 
).
Thank you again for testing this and giving me such valuable feedback! 
This really helps me make the necessary improvements, so I truly appreciate it. Also, where can I find this quiz? It would be super helpful to test the bot further!
