To me, this sounds like:

If AI was good at a certain task then it was a bad task in the first place.

Which is just run of the mill dogmatic thinking.