Why do my coefficients become unbelievably large when I include an interaction term in glm? (R)

Question

Thank you ahead of time, I am somewhat new to R. I am running a model that is trying to find the relationship between policies and public opinion. I am trying to predict whe