That's odd because when I gave ChatGPT my icebreaker interview question that I used a lot at Google it fell right on its face. And this is a question I expect teenagers to ace in one minute.
Intelligence in general, not just of the artificial kind, often comes with blind spots of one kind or another.
ChatGPT isn't even trying to replicate the structure of a human brain, so that it's failure modes are different to ours should not be surprising.
That it can even do this well given it has a network with a similar number of parameters as a (naïve estimate of a) rat's brain, is what's remarkable.