So far as I can explain to, the generative textual content adjusted for every single query. Even so, our evaluate was looking for significant variations within the generative text which might replicate that Google experienced shifted the intent of the first query a bit to return different generative