I'm pretty sure there are plenty of folks at Google who wish they never published the transformer paper.

This is also in line with my subjective experience as someone reading quite a few papers in the field: """The preprint also found that firms based in China consistently published more papers than their counterparts based in the United States."""