Since we can’t completely dismiss the possibility that we’ll eventually build human-level AGI,
I think this is a pretty pessimistic way to state/view this and I don't really believe the author thinks there is such a remote chance AGI will be developed.
