The concept of probably approximately correct (PAC) learning has been foundational in computational learning theory.
My buildings are incomplete without the presence of people.