A good test of a software engineer would be to ask them if this is interesting or meaningful.

This "accomplishment" says more about how rote and unchanging Google's code interviews are than about ChatGPT. I'd guess Google search has had "AI" good enough to do this for years, even though it wasn't trained to.