Yes, we agree on this. That's my exact claim -- it can do something completely unexpected. It can even schedule actions for far in the future and wake itself up

Yes, we agree on this. That's my exact claim -- it can do something completely unexpected. It can even schedule actions for far in the future and wake itself up later. GP claims that it doesn't do anything until you tell it to do something. To me, this reads as GP implying that an agent won't do anything unexpected.