Questions of the form: "AI did this. Is it new/notable/very difficult/ ... ?"

I was the user who posted that question.

I appreciate the discussion here on Meta, but for me personally the answers to my Q have been extremely helpful, and I also really appreciated the follow-up discussion in What is an appropriate role for LLMs in early mathematical research training? -- I spent hours reading the discussion and various linked material.

To give a bit of a background here. I formulated this functional minimization problem (could not solve it), searched the existing literature (did not find anything), asked a couple of more mathematically-versed colleagues (they could not solve it), the colleagues told me that they asked the then-latest ChatGPT as well (and it could not solve it).

At that point I was stuck. And decided that the problem is likely unsolvable in closed-form. That was clearly a mistake. What I arguably should have done at that point, was to post my problem as a question on MathOverflow. However, I have never posted anything on MO so I was feeling a bit shy, and also a bit awkward to ask people here to please solve my research problem.

Six months later, my colleague asked Claude 4.6 and got the derivation and the answer. I was very impressed by that. We (mostly my colleague) spent maybe a week going over the derivation and making sure it is correct, asking Claude to explicitly verify the KKT conditions, catching some hallucinations in its output responses, fixing everything, writing it up in all detail, and finally being satisfied with the derivation. We added all of that to our paper (with detailed disclosure about Claude usage).

I was telling this story -- of how Claude produced this derivation -- to several people. Everybody was impressed. It was a good story. But I had a nagging feeling that maybe the problem was actually not that hard, I just did not reach out to the right people. So eventually, and post factum, I decided to post my MO question.

Note that at that point it would have been disingenuous to ask a MO question about how to solve this minimization problem. I knew how to solve it! Would MO community find my question better, if I actually faked ignorance and asked how to solve it?