My model:

log(y)=β0 + β1x1 + β2log(x2) + β3x1log(x2).

The x1 variable is endogenous and I instrument it with a Z variable. Values of x1 are standardized to be between 0 and 1, where 1 is the highest observed value, and 0 is the lowest.
My second stage estimation gives me these coefficients:
β1=-41.12 β2=0.02 and β3= 3.18

How can I interpret the effect of x1 on y?