Aren't the LLMs also pretty decent code reviewers? Sure, not perfect, but the suggestion that we should use code review as the replacement for the coding challenge 'because AI' seems a bit misplaced...