The team envisions comparing model outputs w/real-world polls, expanding studies to other democracies, examining newer reasoning or "thinking" models, & exploring downstream effects on users. For now, the dataset helps us understand how LLMs are shaping modern political

→ View original post on X — @mit_csail