I'm also trying to understand the implications of this work. Does it imply that some for some functions F(x) = y, you can compute x given the value of y withou

I'm also trying to understand the implications of this work. Does it imply that some for some functions F(x) = y, you can compute x given the value of y without computing the inverse of F ? If so, what constraints does F need to meet for this ?