If you want to keep the point distribution truly random, then as Laurent says, this is impossible.
I’m guessing though that what you’re really after is something not completely regular, but with less variation in spacing than a random distribution.
One way to do this is to iteratively move the points to the centroids of their cells.
It won’t give identical volumes, but they will become more similar than with a random distribution.
lloyds2d.gh (9.7 KB)
naughtyLoop.gh (13.8 KB)
This dirty double data dam trick (gets round GH’s restriction on cyclic data streams without plugins) can be fun to play with, but isn’t really recommended for anything serious, and is likely to cause crashes - use at your own risk!