It can take a while to fully form a position on something. A year isn't really enough for most people to see how deep the rabbit hole goes (unless you were that one CFO that OpenAI had that left after a year). Regardless, publishing a piece like this against a massive company is always a gigantic risk.

The author mentioned that things have become different in the last six months. People are allowed to change their minds when circumstances change.

Doesn’t mean the concerns aren’t valid

I remember when they said GPT-2 was too dangerous to release...