Does the Structure of an algorithm matter for AI Risk and/or consciousness? — LessWrong