Perhaps. 2x is still 2x. And new tools still need to be vetted and learned.

It's strange that the goalpost seems to have moved from "AI is net negative to productivity" to "only 2x improvement isn't worth it"