We shouldn't be surprised when this sort of thing happens in the age of LLMs.

All LLMs do is select the next word based on their training data then repeat.

When people rely on vibecoding, a non-zero fraction of users will inevitably end up with the same results.