Does AGI imply a model will demonstrate morality? Will it produce white-lies when it’s beneficial to it and reject flat out lying when it knows it will get caught or harm others? Will it resolutely stick to a position despite it being a losing one?