The term "friendliness", when used of an AI, isn't intended to indicate friendliness in the same sense we mean it of people. It means that the entity isn't blind to our interests. Notice that I didn't say that the entity has our interests at heart, or that they are its highest priority goal. Those might require intelligence with a human shape. But an SI that was ignorant or uncaring of our interests could do us enormous damage without intending it.