What about algorithmic laws. People complain about affirmative action, because it favors one group of people over another. What if you algorithmically defined

What about algorithmic laws. People complain about affirmative action, because it favors one group of people over another. What if you algorithmically defined quotas instead of saying group X you say of the n groups that are underrepresented in the subset population relative to their existence in the larger population of the country they are weighed favorably by a multiplier of X. This way you can get around writing policies that expressly prefer one group over another. Instead of saying we prefer black people over Asian people into our institution, we prefer "women" over men. The policy should be written as we prefer the lower X% represented ethnic categories to ensure preventing homogeneity and ensure a level of diversity. Of course the issue with this becomes how do you define group identity. People can be divided into infinitesimal categorizations, and how do you define what qualifies as being part of a group. Is Barack Obama black or white? What percentage of genetic markers, indicate that someone deserves a specific label or categorization. Why is that percentage accurate? Why do we prefer one class of discriminated people over another class of discriminated people? The questions go on and on.