Elicitation for Modeling Transformative AI Risks — AI Alignment Forum