Conditioning Predictive Models: Interactions with other approaches — AI Alignment Forum