[delusional]

> Maybe there is one, but it doesn’t support the underlying “and that must mean AI bad” hypothesis as much as the author may think.

It's a tweet. Do you expect thorough null-hypothesis validation from a tweet?